Human Factors Design Guidance for Level 2 and Level 3 Automated Driving Concepts

DOT HS 812 555 August 2018

Human Factors Design

Guidance for Level 2

And Level 3

Automated Driving

Concepts

DISCLAIMER

This publication is distributed by the U.S. Department of Transportation, National

Highway Traffic Safety Administration, in the interest of information exchange. The

opinions, findings and conclusions expressed in this publication are those of the

authors and not necessarily those of the Department of Transportation or the

National Highway Traffic Safety Administration. The United States Government

assumes no liability for its content or use thereof. If trade or manufacturers’ names

or products are mentioned, it is because they are considered essential to the object

of the publication and should not be construed as an endorsement. The United States

Government does not endorse products or manufacturers.

Suggested APA Format Citation:

Campbell, J. L., Brown, J. L., Graving, J. S., Richard, C. M., Lichty, M. G., Bacon, L. P., … & Sanquist,

T. (2018, August). Human factors

design guidance for level 2 and level 3 automated driving

concepts (Report No. DOT HS 812 555). Washington, DC: National Highway Traffic Safety

Administration.

REPORT DOCUMENTATION PAGE

Form Approved

OMB No. 0704-0188

1. AGENCY USE ONLY (Leave blank) 2. REPORT DATE

August 2018

EPORT TYPE AND DAT

OVERED

4. TITLE AND SUBTITLE

Human Factors Design Guidance for Level 2 and Level 3 Automated Driving Concepts

5. FUNDING NUMBERS

Contract No. DTN22-11-00236/0011

6. AUTHORS

John L. Campbell, James L. Brown. Justin S. Graving, Christian M. Richard, Monica G.

Lichty, L. Paige Bacon, Justin F. Morgan, Hong Li, Diane N. Williams, Thomas Sanquist

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS

Battelle Memorial Institute

505 King Avenue

Columbus, Ohio 43201-2696

8. PERFORMING ORGANIZATION

EPORT NUMBER

9. SPONSORING/MONITORING AGENCY NAME AND ADDRESS

National Highway Traffic Safety Administration

1200 New Jersey Avenue SE.

Washington, DC 20590

10. SPONSORING/MONITORING

AGENCY REPORT NUMBER

DOT HS 812 555

11. SUPPLEMENTARY NOTES

Dr. Paul Rau was NHTSA’s Contracting Officer’s Representatives (COR) for this effort.

12a. DISTRIBUTION/AVAILABILITY STATEMENT

This document is available to the public from the National Technical Information Service,

www.ntis.gov.

12b. DISTRIBUTION CODE

13. ABSTRACT

High automation presents the opportunity to increase the safety, mobility, and efficiency of the existing road network, and has been

anticipated for decades. It was not, however, until the development of sophisticated sensing and computing systems that such vehicles

became technically feasible. Many automakers and Tier-1 suppliers are developing or testing vehicles with some form of automation.

In support of the motor vehicle automation effort, NHTSA is planning a automation driving system (ADS) research program in

coordination with other USDOT agencies with the goal of improving motor vehicle safety. Driver vehicle interface (DVI) design

guidance has been developed as part of a larger research effort intended to perform an initial human factors assessment of driver

performance and behavior under Level 2 and Level 3 automated driving. Safe and efficient operation of any motor vehicle requires

that the DVI be designed in a manner consistent with driver limitations, capabilities, and expectations. This document is intended to

assist DVI developers achieve these outcomes.

14. SUBJECT TERMS

Automation, Automated Driving Concepts, Levels of Automation, Driver Vehicle Interface, DVI,

Driver Performance

15. NUMBER OF PAGES

127

16. PRICE CODE

17. SECURITY CLASSIFICATION OF

REPORT

Unclassified

18. SECURITY CLASSIFICATION OF

THIS PAGE

Unclassified

19. SECURITY

CLASSIFICATION

OF ABSTRACT

20. LIM

ITATION OF

ABSTRACT

Table of Contents

Chapter 1. Introduction ............................................................................................................. 1

Background ............................................................................................................................ 1

Overview of Automation ......................................................................................................... 2

Automation in Vehicles ........................................................................................................... 4

General Design Issues for Automated Driving Systems ........................................................ 6

Current and Future Directions of Vehicle Automation ............................................................ 8

Current Best Practices ......................................................................................................... 10

Driving Automation Research and Design Guidance Development ..................................... 12

Scope of this Document ....................................................................................................... 13

Limitations of This Document ............................................................................................... 13

Organization of This Document ............................................................................................ 13

Chapter 2. How to Use This Document ................................................................................. 15

Two-Section Format ............................................................................................................. 15

The First Section .................................................................................................................. 15

The Second Section ............................................................................................................. 16

Selection of Font Sizes in This Document ............................................................................ 17

Chapter 3. General DVI Design Guidance for Level 2 and Level 3 Automation ................. 18

Current Automation Mode and Status .................................................................................. 18

Suitable Display Properties for Automation Mode and Status Messages ............................ 21

Communicating Transfer of Control From Driver to System ................................................ 24

Communicating Transfer of Control From System to Driver ................................................ 27

Developing and Maintaining Driver Mental Models .............................................................. 30

Special Considerations for Level 2 and Level 3 Automation ................................................ 33

Developing Driver Training Material for Automated Driving System Applications ................ 35

Incorporating Etiquette Into the Design of Automated Systems ........................................... 38

Chapter 4. Message Characteristics ...................................................................................... 40

Designing Messages for Driver Comprehension .................................................................. 41

Message Complexity ............................................................................................................ 43

Selection of Sensory Modality .............................................................................................. 45

Multimodal Messages ........................................................................................................... 47

Chapter 5. Visual Interfaces ................................................................................................... 49

Locating a Visual Display ..................................................................................................... 50

Display Glare ........................................................................................................................ 51

Head-Up Displays ................................................................................................................ 54

Using Color ........................................................................................................................... 56

Selecting Character Height for Icons and Text ..................................................................... 58

Temporal Characteristics of Visual Displays ........................................................................ 60

iii

Chapter 6. Auditory Interfaces ............................................................................................... 62

Perceived Urgency of Auditory Warnings ............................................................................ 63

Perceived Annoyance of Auditory Warnings ........................................................................ 65

Loudness of Auditory Warning Signals ................................................................................ 67

Chapter 7. Haptic Interfaces .................................................................................................. 70

Selecting a Haptic Display ................................................................................................... 71

Improving Distinctiveness of Haptic Displays ...................................................................... 73

Accommodating for Vibrotactile Sensitivity Across the Body ............................................... 75

Generating a Detectable Signal in a Vibrotactile Seat ......................................................... 77

Chapter 8. Driver Inputs ......................................................................................................... 80

General Guidance for Driver-DVI Interactions ..................................................................... 81

Control Placement ............................................................................................................... 83

Voice Recognition Inputs ..................................................................................................... 85

Chapter 9. Glossary ................................................................................................................ 88

Chapter 10. Index ...................................................................................................................... 95

Chapter 11. Abbreviations ....................................................................................................... 97

Chapter 12. Equations .............................................................................................................. 99

Chapter 13. Relevant Documents From the United States Department of

Transportation, SAE International, and International Organization for

Standardization ....................................................................................................................... 101

Chapter 14. References .......................................................................................................... 108

Chapter 1. Introduction

Background

Motor vehicle automation can potentially improve highway safety by supporting or

supplementing the driver, thereby providing precise vehicle control during normal driving, and by

maintaining appropriate driver attention to traffic and roadway conditions. Although it is expected

that automated systems will not have universal capabilities in all traffic and environmental

conditions for some time to come, applications in motor vehicles will likely include a driving

experience of seamless transitions between automation and manual control of motion control

system functions in complex and rapidly changing conditions.

Higher levels of driving automation systems present the opportunity to greatly increase the safety,

mobility, and efficiency of the existing road network. Automated vehicles, however, have been

“coming soon” since the first half of the 20th century, when vehicles guided automatically along a

highway were described at the 1939 World’s Fair (O’Toole, 2009). It was not until the

development of sophisticated sensing and computing systems that such vehicles became

technically feasible. Prior large-scale efforts, such as the Federal Highway Administration

(FHWA) Automated Highway System (AHS) research, provided some information as to the

potential of automated systems. Yet it was not until newer explorations of ground vehicle

automation and highly-visible events such as the 2004 and 2005 Defense Advanced Research

Projects Administration (DARPA) Grand Challenge that the near-term potential of vehicle

automation became apparently to the broader community.

While some relevant research exists from AHS or adaptive cruise control (ACC) projects, current

and near-term implementations of driving automation have not been extensively researched.

Further, at the time of writing this document, a number of automakers and Tier-1 suppliers are

currently developing or testing vehicles with some form of automation. Thus, constructing

appropriate design guidance for automated driving requires, and must be partially based upon an,

understanding the broader field of human factors in automation and human-automation

interaction, as well as attempting to understand how automation is likely to be implemented in the

near-future.

In support of the motor vehicle automation effort, the National Highway Traffic Safety

Administration’s Office of Crash Avoidance and Electronics Systems Safety Research is planning

an automated systems research program in coordination with other USDOT agencies including

the Research and Innovative Technology Administration (RITA),

FHWA, Federal Motor Carrier

Safety Administration (FMCSA), and Federal Transit Administration (FTA). The goal of the

program is to improve motor vehicle safety by defining the requirements for automation in

driving that is: (1) functionally safe and electronically reliable; (2) operationally intuitive for

The Research and Innovative Technology Administration (RITA) was moved by Congress in 2014 to the Office of the Assistant

Secretary for Research and Technology (OST-R), a part of the Office of the Secretary of Transportation (OST), which is home to

all the program offices and statistics and research activities previously administered by RITA.

drivers under diverse driving conditions; (3) compatible with driver abilities and expectations; (4)

supportive of improving safety by reducing driver error; (5) operational only to the extent granted

by the driver and always deferent to the driver; and, (6) secure from malicious external control

and tampering. Addressing the human factors questions is central for accomplishing these goals.

A key element of vehicle automation is the driver-vehicle interface or DVI. The DVI refers to

vehicular displays that present information to the driver, and controls that facilitate the driver’s

control of the vehicle as a whole as well as the status of various vehicle components and

subsystems. Safe and efficient operation of any motor vehicle requires that the DVI be designed

in a manner that is consistent with driver limitations, capabilities, and expectations. This

document is intended to assist DVI developers to achieve these outcomes. This DVI design

guidance has been developed as part of a larger research effort—Human Factors Evaluation of

Level 2 and Level 3 Automated Driving Concepts—that is intended to perform an initial human

factors assessment of driver performance and behavior under Level 2 (L2) and Level 3 (L3)

automated driving.

Overview of Automation

Uses of Automation

Automation has been defined as a device or system that accomplishes (partially or fully) a

function that was previously or conceivably could be performed (partially or fully) by a human

operator (Parasuraman, Sheridan, & Wickens, 2000). Historically, the use of automation has been

found in process control and aviation but, today, examples of automation can be found as well in

non-industrial and personal uses. Automation in vehicles has the potential to help drivers who

choose to engage in distracting behaviors (e.g., text messaging while driving) or who are

experiencing a high level of workload by filling the gap between the driving demands and the

capabilities of the driver. While there are many automated systems in the vehicle (such as

automated driver assistance systems or safety systems that intervene in the absence of driver

responses to specific situations), for the purpose of this document automation is a system that

physically performs a specific combination of driving functions for the driver (e.g., maintain

headway and steer the vehicle in a lane during a traffic jam, choose and execute a route to a

chosen destination, or park the vehicle).

Philosophies of Automation

In developing a system that will include some degree of automation, system planners and

designers may begin by first determining which functions the automation will take over from the

driver and which functions the driver will continue to perform. Research in human-automation

interaction provides four general automation philosophies that provide a way to view, or in some

cases determine, this allocation of functions between the human and the automation:

• The “left-over” or residual function principle is the earliest automation philosophy. Under

this philosophy, the automation is designed to perform as many functions as possible, with

the remaining functions being allocated to be performed by the human. The rationale

behind this philosophy is that since the automation can be designed to perform functions

or tasks more quickly, reliably, and with fewer errors than a human, it should perform as

many of the tasks as possible. This philosophy proposes that only functions that can be

automated completely and will not suddenly require the intervention and support of a

human should be automated (Hollnagel & Bye, 2000).

• The compensatory principle, or “Fitts List” (Fitts, 1951), is a list or table of the strong and

weak features of humans and machines used as a basis for assigning functions and

responsibilities to the various system components. As shown in Table 1-1, the function

that humans are better at are the functions that would be assigned to the human to perform

and functions that machines are better at are those that would be assigned to the

automation to perform. Under this automation philosophy, humans are seen as mainly

responding to what happens around them and their actions are the result of processing

input information using whatever knowledge they may have (i.e., their mental models;

Hollnagel, 2004).

Table 1-1. Human and machine function allocation.

Humans are better at: Machines are better at:

• Detecting small amounts of visual or acoustic energy

• Perceiving patterns of light or sound

• Improvising and using flexible procedures

• Storing large amounts of information for long periods

of time and recalling relevant facts at the appropriate

time

• Reasoning inductively

• Exercising judgment

• Responding quickly to control signals and applying

great force smoothly and precisely

• Performing repetitive, routine tasks

• Storing information briefly and then erasing it

completely

• Reasoning deductively, including computational

ability

• Handling highly complex operations, i.e., doing

many different things at once

• Dynamic function allocation (as opposed to the static approaches above) is a

complementary approach that enables the human and automation to trade off which

functions each performs based on the current situation. Instead of focusing on what types

of functions the automation is better at performing and what types of functions the human

is better at performing, as seen with the Fitts List, the focus is now on how humans and

the automation can complement and support each other to achieve the overall purpose

(Grote, Weik, Wäfler, & Zӧlch, 1995; Wäfler, Grote, Windischer, & Ryser, 2003). This

approach aims to sustain and strengthen the human’s ability to perform efficiently by

focusing on the work system in the long term, including how routines and practices may

change because of learning and familiarization.

• Adaptive function allocation is an extension of the complementary approach which

assumes criteria to determine whether functions must be reallocated, how and when based

on changes in the operating environment, loads or demands to operators and performance

of operators (Inagaki, 2003). An automation system that operates under an adaptive

function allocation is called adaptive automation. Adaptive automation can be used to help

regulate workload by having the operator control a process during periods of moderate

workload and then hand-off control of particular tasks when workload either rises above

or falls below an optimal level (Hilburn, Molloy, Wong, & Parasuraman, 1993;

Parasuraman & Wickens, 2008). Adaptive automation can also assist in keeping the

operator in-the-loop by altering the level of automation being used. This information

should be used with caution, however, as frequent cycling between automated and manual

control may sometimes, but not always, cause a decrease in performance.

Automation in Vehicles

NHTSA’s Preliminary Statement of Policy Concerning Automated Vehicles (NHTSA, 2013)

provided an initial taxonomy of road vehicle automation that included five levels of automation.

Additionally, the Statement of Policy provided information on the developments in automated

driving at the time and an overview of NHTSA’s automated systems research program. The SAE

also defined a taxonomy (SAE J3016, 2014) that consists of six levels of automation, which was

adopted by NHTSA (described below in Table 1-2).

Table 1-2. Summary of SAE International driving automation levels.

Level and Name Description

Level 0 (L0)

No Driving Automation

The human driver does all the driving.

Level 1 (L1)

Driver Assistance

Vehicle is controlled by the driver, but some driving assist features may be included

that can assist the human driver with either steering or braking/accelerating, but not

both simultaneously.

Level 2 (L2)

Partial Driving

Automation

Vehicle has combined automated functions, like speed control and steering

simultaneously, but the driver must remain engaged with the driving task and monitor

the environment at all times.

Level 3 (L3)

Conditional Driving

Automation

An automated driving system on the vehicle can itself perform all aspects of the

driving task under some circumstances. Driver is still a necessity, but is not required

to monitor the environment when the system is engaged. The driver is expected to be

takeover-ready to take control of the vehicle at all times with notice.

Level 4 (L4)

High Driving

Automation

The vehicle can perform all driving functions under certain conditions. A user may

have the option to control the vehicle.

Level 5 (L5)

Full Driving Automation

The vehicle can perform all driving functions under all conditions. The human

occupants never need to be involved in the driving task.

The SAE International taxonomy is applicable to all implementations of driving automation, is

technology agnostic, and has some important distinctions, assumptions, and implications. From

these, functional distinctions and assumptions follow the associated role of the driver of the

vehicle at each level.

L2 vehicles are interesting as they may, in some implementations, utilize existing production

technologies, such as radar and machine vision technology, to provide robust automation. L3

vehicles may utilize these or a completely different set of technologies in support of automation.

As technically defined in SAE J3016.

From the average driver’s perspective, the actions of a highly-performing L2 system and any L3

system performing the same automated function may appear to be the same. The highway speed

automation in L2 or L3 controls the system by keeping the vehicle within its lane and controlling

the speed and headway to any leading vehicles. Ensuring the driver builds an appropriate mental

model, the driver is provided with information about the status of the automation, and is aware of

what the automation may or may not do while actively controlling the vehicle may yield

significant benefits. Yet there is a lack of published research examining how drivers form mental

models of different levels of vehicle automation, and how those mental models are applied (and

especially how mental models are applied between different levels of vehicle automation). While

research into driver performance with L2 and L3 systems is in a nascent stage, examining drivers’

interactions with different levels of automation may provide highly useful information for DVI

design guidance in such vehicles.

As the level of automation increases, the driving role shifts from the driver to the vehicle. Under

L2 automation, drivers are expected to serve as a monitor of automation as they are ultimately

trusted to ensure the safe operation of the vehicle. The SAE definition states that L2 automated

driving systems can release control with little or no advance warning. At higher levels of

automation, the vehicle starts assuming more aspects of the driving role while under automated

control. In contrast to L2, under L3 automation drivers are not expected to monitor the

environment when the system is engaged, but they are expected to be takeover-ready to take

control of the vehicle at all times with notice. Therefore, the L2 driver is expected to be alert and

monitoring the road continuously. The L3 driver is not expected to be paying attention to the road

at all times, but is expected to be takeover-ready with advance notice.

This is a unique situation in terms of the driver’s attention. The L2 driver has been relieved of the

control level of driving task, and potentially some of the maneuvering level (Michon, 1985).

Thus, the driver of the L2 vehicle has transitioned from the role of active driver to one of a

monitor of driving automation. Under L2 automation, the driver is expected to monitor the road,

the performance of the automation, and be ready to intervene if something functions incorrectly or

if asked to do so. In contrast, the L3 driver has transferred control, maneuvering, and perhaps

some of the strategic choices to the automation. The driver must be ready to resume control at any

time with advance notice in L3 automation. The L3 driver, largely relieved of the role of driving,

is not required to constantly monitor the vehicle or road status; the L3 driver is only asked to be

ready to intervene if warned in advance.

Note that the L2 driver is engaged in the monitoring task even as event rate and workload are

reduced to what may be described as a vigilance task (O’Hanlon & Kelley, 1977). While

monotonous driving over time has been identified with poorer performance in both laboratory

evaluations of driving and naturalistic evaluations of long-duration commercial vehicle driving

(Thiffault & Bergeron, 2003; Soccolich et al., 2013), it is currently unknown how drivers will

perform in terms of sustained attention under longer duration L2 or L3 automated driving. Related

to this, emerging research suggests that drivers, when relieved of actively controlling the vehicle,

may engage in a variety of non-driving tasks that can involve significant levels of distraction

(Llaneras, Salinger, & Green, 2013). Taken together, the potential for a vigilance-like state and

behaviors that would be termed distraction in manual driving provide the foundation for important

guidance on L2 and L3 automation DVIs. Understanding lessons learned in how to provide both

safety-critical and non-safety critical messages to operators who are either in a vigilance state or

involved in secondary (divided attention) tasks provides useful information for the designer of L2

and L3 automated system DVIs. The implication for design guidance is that, under L2 and L3

operations, automation-specific messages (and especially safety-critical messages) may need to be

provided in a distinctive and highly salient manner as the driver may not be currently monitoring

the forward roadway (Llaneras et al., 2013). Further research in this area is likely to provide

highly useful guidance and understanding best practices.

The definitions of L2 and L3 automated systems also have implications for the DVI in terms of

transfer times. The L2 system driver is expected to be able to take control with little to no advance

warning. The L3 system’s driver is expected to be able to take control with a sufficient and

comfortable transition time. The issue of advance-notice transfer time may be considered both in

terms of the active status of the road (the origin of the take control message) and of the driver’s

status (e.g., engaged in another task for the last 10 minutes). Perhaps more critical from a DVI

design standpoint is the “no advance warning situation” possible in L2 vehicles. L2 vehicles have

the potential to fail in such a manner that no advance warning is provided (a “silent failure”). One

such example of this would be that of a vehicle incorrectly following lane markings and

(intentionally) drifting onto the shoulder instead of continuing on the road. In this hypothetical

example, the inattentive driver would not be informed of this until the vehicle encountered the

infrastructure-based warning of the rumble strip. The issues surrounding no advance warning

transfers and how drivers sample the road in different levels of automated driving, as well as

transition times for both L2 and L3 automated systems, remains largely unexplored. The need for

research in this topic was acknowledged, however, and existing knowledge considered in the

design principle generation process. Some of the applicable lessons learned from the broader

human-automation interaction field are described in the following section.

General Design Issues for Automated Driving Systems

Due to the fact that automation has been used in a variety of domains, for decades in some cases,

a large body of research exists regarding what can be considered basic design issues when

developing automated systems. These issues are listed here.

• Trust in automation (Lee & Moray, 1994; Muir & Moray, 1996; Lee & See, 2004;

Rajaonah, Anceaux, & Vienne, 2006)

As vehicle automation is able to perform more functions for the driver, the driver’s

confidence in the automation’s ability to perform these functions and a willingness to rely

on information provided by the automated system becomes more important. For example,

if the automation system alerts the driver to a hazard, even when the hazard is not

immediately observable, the driver prepares to take the necessary action to prevent from

hitting or running over the hazard. In terms of L3 automation, in which the driver can cede

full control of all safety-related functions to the automation under certain traffic or

environmental conditions, the topic of trust in automation is particularly relevant due to

the fact that the automation is the primary source of ensuring safe operation of the vehicle.

• Misuse, disuse, and abuse of automation (Parasuraman & Riley, 1997)

Using automation when it should not be used, ignoring or turning off automated alarms or

safety systems due to reoccurring false alarms, or implementing automation without

regard for the consequences for human performance can affect a driver’s trust in the

automation system. Proper design of the automation system and appropriately calibrated

trust in the automation system may help support drivers’ appropriate use of the

automation. As higher levels of automation are implemented (L2 and L3), this is a

particularly relevant topic because the consequences to safety are higher as the driver is

less engaged in driving (in L2 automation) or is performing fewer safety-related functions

(in L3 automation).

• The out-of-the-loop problem (Endsley & Kiris, 1994; Stanton & Young, 1998; Seppelt &

Lee, 2007)

If a driver is not provided with the proper information about the automation, and

potentially road/traffic status, he or she may suffer from a diminished ability to detect

automation failures and to re-claim manual control of the vehicle if necessary. This is a

topic that is becoming more important as higher levels of automation (L2 and L3) are

implemented in vehicles and drivers are potentially able to perform non-driving related

tasks while in the vehicle. At these levels of automation, the automation is performing

nearly all, if not all, of the primary driving tasks, meaning that a driver could be so

involved in a non-driving related task (e.g., texting, checking e-mail, reading a book, etc.)

that he or she could completely miss a warning or message from the automation, possibly

leading to a crash or other incident.

• Failures in automation (Endsley & Kiris, 1994; Stanton & Young, 2000)

As automation systems have been implemented in various domains—including in

vehicles—the reliability of these systems has improved. Even with the most advanced

automation available, however, the reliability will never be perfect. Performance after a

failure may be better if the driver has a realistic understanding of the reliability of the

system, and the types of system failures that could potentially occur. In a scenario in

which the automation fails, there may be benefits if the driver could be alerted to the

failure with plenty of time to re-claim manual control of the vehicle from the automation.

As L2 and L3 automation are designed and implemented in vehicles, the status of the

automation and being alerted to a failure of the automation system may become even more

important due to the fact that a driver is unlikely to be in contact with either the steering

wheel or the pedals. This means that the sooner he or she can tell that the automation has

failed or is alerted to the fact that the automation as failed, the sooner he or she can

prepare to and ultimately re-claim manual control of the vehicle.

• Workload and the implementation of automation (Young & Stanton, 1997)

The implementation of automation is often expected to reduce driver workload, which

during periods of high workload can be helpful. However, underload—resulting from

fewer driving tasks—can also occur, potentially causing a driver to be become bored due

to the lack of tasks to perform. The prevention of the driver becoming bored or fatigued

due to underload is particularly relevant in terms of (primarily) L2 automation since the

reduction of driver-performed functions is great to begin with. Therefore, L2 automation

could be adaptive, meaning that there could be a balance between which functions the

driver was still performing and which functions the automation was now performing,

impacting the workload experienced by the driver.

• Clumsy automation (Wiener, 1989)

If the automation is designed to perform the easier task for the driver, leaving him or her

to perform the more challenging task, the result is clumsy automation. In terms of

workload, with this type of automation, the automation reduces the workload experienced

when it is already low and increases the workload experienced when it is already high.

This topic is important because if automation makes the driving situation more

challenging than it already is, it defeats the purpose of implementing automation. When

designing L2 automation, designers and systems planners could consider adaptive

automation to address this concern. As mentioned above, this may help keep the workload

experienced by the driver balanced and will potentially reduce the occurrence of clumsy

automation altogether.

• Mode awareness (Sarter & Woods, 1995)

When a driver is using a level of automation and fails to detect a change in the level of

automation, it is called a mode awareness failure, or automation surprise, e.g., when a

driver engages a lane centering system and then, for an unknown reason, the lane

centering system disengages. Without proper feedback, the driver may not know the lane

centering system is disengaged and might not take the necessary steps to re-claim manual

control of the vehicle and to correct or change the course of the vehicle. As the ability of

the automation to perform functions for the driver increases, the potential for the driver to

possibly engage in non-driving related activities also increases, making the process of

alerting the driver to any critical changes in automation mode more important.

Current and Future Directions of Vehicle Automation

Available Technology

Currently, L1 automation technologies are available in some, but not all, new light vehicles on the

market today and L2-type technologies are beginning to become available from different

manufacturers (NHTSA, 2013). L1 automation technologies are designed to assume a portion of

driving authority over a primary control (e.g., adaptive cruise control).

L2 technologies could allow the driver to physically (but not cognitively) disengage from the

driving task and assume the role of monitoring the road and vehicle performance. Some L2

technologies, such as ACC plus lane centering, are also available in new light vehicles on the

market.

L3 technologies allow the driver to cede both the monitoring and control role for the full driving

task to the automation under certain conditions but expect the driver to be takeover-ready when

the system may request it. An example of an L3 concept is traffic jam pilot, which allows drivers

to engage the automation and cease monitoring the road. As L2 technologies allow the driver to

physically disengage from the driving task, L3 technologies allow the driver to relegate full

control of the driving task to the vehicle. These L3 technologies, while experiencing varying

degrees of maturation while in development, are not currently available in the United States light

vehicle market, yet they are expected to be available in the near future.

L4 and L5 technologies, which are expected to only require the driver to input a destination while

the automation monitors the roadway and performs all safety-critical functions, do not yet exist

outside of advanced research concepts. L4 systems are constrained in a domain, and L5 systems

can operate without such limitations.

Role of the Driver

With the implementation of automation in the vehicle come changes in the role of the driver in the

driving task. When performing the driving task manually (under L0 or L1 automation), the driver

is actively participating in the task. The driver is completely responsible for control of the vehicle

and monitoring the roadway for safe operation of the vehicle. When a level of automation is

added, however, part of the task is taken over by the automation, leaving the driver to participate

in less of an active manner and more of a passive manner during normal operating conditions. For

example, under L1 automation the driver may still be responsible for steering the vehicle, but is

not responsible for maintaining a safe following distance behind the vehicle in front of him or her

when using ACC. Under L2 automation (extending the previous example to include both ACC

and lane centering) the driver is expected to monitor the performance of the vehicle continuously,

but the system may manage vehicle headway and lane position appropriately for extended period

of time without intervention from the driver.

Increased automation and the accompanying shift to a role of automation monitor presents the

potential for driver underload. Research from domains such as unmanned vehicles and control

systems has suggested that increased time in a monitoring role can lead to reduced performance

(Cummings, Mastracchio, Thornburg, & Mkrtchyan, 2013). It is unclear, however, whether this

can be extended to driving. As automation technologies improve and evolve, the role of the driver

in the driving task must be assessed and reassessed as changes continue.

Automation Transitions

As higher levels of automation are researched, designed, tested, and eventually implemented,

additional design considerations will need to be discussed. A main concern with near-term and

future levels of automation is the occurrence of planned and unplanned transitions between the

levels of automation and the impact that these transitions will have on the timing of delivering

information to the driver (e.g., much before the transition, near the occurrence of the transition),

the modality of delivering the information to the driver (e.g., for the auditory modality, a single

tone, a consistent tone, an auditory icon), and what specific information that will be

communicated to the driver (e.g., what function the automation was performing but is not now,

that the automation went from L3 to L1). To keep the driver in-the-loop as best as possible,

enabling him or her to re-claim manual control of the vehicle more quickly if necessary,

addressing the following questions may be of value: what information should drivers receive

about the transition in level of automation, when should drivers be alerted about the transition

between one level of automation and another level of automation, and how should drivers be

informed about the transition in level of automation? This is a particularly important topic due to

the fact that near-term in-vehicle automation (L2 and L3) will, at least in the beginning, likely

consist of short cycles between the different levels of automation, meaning a short duration of

time between engaging the automation and then returning to manual mode.

Short-cycles of automation can also be found in the aviation domain. In a study by Parasuraman,

Hilburn, Molloy, and Singh (1991), participants monitored three flight-related functions that

could be automated or performed manually, with a shift from manual to automatic control and

back occurring every 10 minutes during the 30-minute session. When the functions were

automated, participants were required to perform a supervisory control task of the automation.

Results from this study showed performance benefits in all three flight functions and no evidence

of costs for the three flight functions. Also, results showed that dynamic automation shifts when

transitioning between levels of automation benefitted the performance of the flight-related tasks,

without evidence of costs to performance following the return to manual control. While this can

be used as a starting point for thinking about short-cycle automation in vehicles, the results cannot

be generalized from aviation to surface vehicles without further research.

Maintenance of Driving Skills for Drivers

Another concern with the implementation of higher levels of automation (L2 and L3) is the

degradation of some driving skills due to reliance on automation and lack of exposure to manual

driving. At the current level of automation this does not seem to be much of an issue; as higher

levels of automation become available, however, it can become much more significant. Already a

recognized issue in aviation, prompting the Federal Aviation Administration (FAA) to release a

safety notice recommending that pilots fly in manual mode more than using the autopilot (FAA,

2013), skill degradation can not only apply to psychomotor skills, but also for decision-making

skills (Miller & Parasuraman, 2007). The maintenance of basic skills is important because, as

previously mentioned, reliability of automation is still not perfect. There is a good chance that

drivers will need to re-claim manual control of the vehicle at some point in time.

Awareness of the System by Drivers

Until L4/L5 technology expecting a driver to only provide destination input and then let the

vehicle take care of all safety-critical functions, including monitoring the roadway exists, there

will continue to be some level of shared authority between the driver and the vehicle over the

tasks that make up driving. For lower levels of automation, design of vehicle automation that

supports the driver’s full understanding of the capabilities and limitations of the automation, as

well as the driver’s awareness of the automation’s current state, could yield safety benefits. The

failure to remain aware of the state of the automation and the overall status of the driving task, if

the automation has failed, could potentially increase the likelihood of collisions.

Current Best Practices

While there is not yet enough relevant research about the current and near-term implementation of

vehicle automation to develop full and specific design requirements, there is enough research to

recognize some best practices that could be considered when designing and implementing vehicle

automation. Table 1-3 below provides a summary of factors that may influence the effectiveness

of an automation design, the best practice to follow regarding that factor, and key references that

can provide more information on the best practice. These best practices can be used by designers

and system planners as a starting point, however, it is still unknown how exactly they will apply

when designing and implementing L2 and L3 automation.

Table 1-3. Summary of factors that can influence the effectiveness of an automation design.

Factors that Influence

Effectiveness

Best Practice Key Reference(s)

Driver Understanding of

Automation

Automation should not degrade a driver’s mental

model of how the automation system functions.

Goodrich & Boer, 1999

Preservation of Situation

Awareness

Drivers should be able to maintain good situation

awareness in case of an automation failure (out-of-

loop familiarity).

Endsley & Kiris, 1995;

Endsley & Kaber, 1999;

Seppelt & Lee, 2007

Automation-Induced Errors

Automation should not introduce new sources of

driver error.

Sarter & Woods, 1995

Appropriate Level of Trust

The level of trust that drivers place in the

automation should be commensurate with the

capabilities of the automation. Drivers should not

become over-reliant on automation and therefore

unable to perform key driving tasks.

Lee & See, 2004;

Muir & Moray, 1996;

Lee & Moray, 1994

Behavioral Adaptation

Automation should not lead drivers to adopt “bad

habits” (i.e., not paying attention to the road) that

can lead to unsafe driving in some situations or in

different vehicles.

Parasuraman & Riley, 1997;

Wilde, 1998;

Rudin-Brown, & Parker, 2004;

Jamson, Merat, Carsten, & Lai, 2013

Allocation of Functions

Allocation of a function between manual and

automated processes should be determined by task

suitability (i.e., Fitts List, adaptive functional

allocation), rather than application/engineering

capabilities.

Hollnagel & Bye, 2000;

Fitts, 1951;

Grote, Weik, Wäfler, & Zölch, 1995;

Inagaki, 2003;

Hilburn, Molloy, Wong, &

Parasuraman, 1993;

Hollnagel, 2004

Definition of the Driver’s

Role

The driver’s role and functions in the automated

application should be clearly apparent.

Hollnagel & Bye, 2000;

Inagaki, 2003

Incomplete Automation

Incomplete or “clumsy” automation should be

avoided (i.e., when easy tasks are automated but

complex ones remain manual) unless it constitutes

an appropriate allocation of function for the

application.

Wiener, 1989

Status Display Availability

Display of system status should be available on

demand (i.e., a persistent visual notification).

Flemisch & Schieben, 2009

Status Display Modality

Status display modality should be consistent with

or complement the type of information presented

(i.e., visual for persistent information, auditory or

conspicuous for change information.

Lee, McGehee, Brown, & Marshall,

2006

Change in Automation

Status

Display of system status should indicate changes in

system status. Status changes in critical tasks

should be adequately conspicuous, but not

disruptive to the driving task.

Hancock, 2007;

Seppelt & Lee, 2007

Driving Automation Research and Design Guidance Development

The best practices summarized above provide the beginnings of a foundation for thinking about

the DVIs of automated driving systems. These are some of the concepts considered during the

process of assembling the design guidance provided in this document. However, this is far from a

comprehensive foundation in terms of supporting the design of actual vehicle automation systems.

Much more research is needed to provide the kind of science-based guidance that could benefit

system developers and designers. In particular, further research about what information should be

presented, how it should be presented (mode, location, format, location, etc.), and when it should

be presented could be very helpful. Understanding how these factors apply across different levels,

and within different implementations of automation are also important. To that end, research

performed so far in the Human Factors Evaluation of L2 and L3 Automated Driving Concepts

project has culminated in this set of human-centric design guidance for the DVI of automated

driving systems. This design guidance represents both the current state of published science, as

well as a beginning step for a rapidly evolving field.

This document may assist manufacturers in minimizing the unintended consequences of motor

vehicle automation and help support designers in creating systems that are compatible with driver

limitations and capabilities. Consistent with this goal, the development team has focused on

providing a clear, relevant, and easy-to-use reference of human factors guidance for in-vehicle

DVI design and operation within L2 and L3 automated driving environments. The development

team has worked cooperatively with other project team members to ensure that relevant research

and suggestions are integrated into the design guidance development process. Overall, the DVI

design guidance is intended to:

• Be concise, clear, and easy to use.

• Include graphics-based design tools and examples that can be used by designers who are

unsophisticated regarding human factors issues and practices.

• Include discussions of critical design issues and special design considerations when, for

example, design trade-offs must be made or design constraints exist.

• Serve as a repository for relevant standards and guidelines.

• Support increased awareness and knowledge of relevant standards, human factors

concepts, and user characteristics among DVI developers and designers.

While DVI design guidance can be a valuable tool and resource for designers, they have their

limitations, and for L2 and L3 ADS, human drivers are ultimately responsible for driving safely.

Also, we recognize and appreciate the complexities and challenges associated with developing

automated driving systems; the design guidance is intended to augment, but not replace, the

judgment and experience of DVI developers.

Due to the variety of data sources used in this document, users may be uncertain regarding the

applicability of individual data sources to safety-related vs. non-safety-related DVI questions. In

general, when considering the applicability of individual design topics to a specific DVI design

question, users of this document should carefully consider the DVI question or issue they are

addressing relative to the characteristics (e.g., objectives, research and analytical methods,

limitations) of the original data sources cited, our syntheses of and conclusions regarding these

data sources.

Scope of This Document

This document provides goals and guidance for the design and development of DVIs for L2 and

L3 automated driving systems.

Limitations of this Document

As noted above, this document reflects an initial step in thinking about the information needs of

drivers in automated driving systems and trying to translate those needs into DVI design

guidance. There is relatively little published research available to support the development of

detailed design specifications for the DVIs of L2 or L3 automated systems. Thus, this document

is far from a comprehensive foundation in terms of supporting the design of actual vehicle

automation systems. In many respects, it highlights key knowledge gaps and points towards what

we do not yet know about the DVIs of automated driving systems. Future research may contribute

to the science-based guidance that system developers and designers may appreciate. In particular,

further research regarding what information should be presented, how it should be presented,

when it should be presented, and what kinds of driver behavior/performance/acceptance

challenges are associated with L2 and L3 systems could be beneficial to the community.

Understanding how these factors apply across L2 and L3 systems (and other levels in the future),

and across different implementations of automation (e.g., systems that provide longitudinal and

lateral control at highway speeds vs. parking assist systems vs. systems that provide drivers

assistance within traffic jams) are important issues for future research.

Organization of This Document

Beyond the Introductory chapter, this document consists of a series of chapters containing DVI

design guidance. Each chapter contains a set of subtopics relevant to a specific design

characteristic or element. Chapter 2 provides an overview of the format and content of these

design-specific chapter topics (Chapters 3 through 8). Following the design chapters are a set of

reference chapters with supplemental information which may be useful for either a specific topic

or for DVI design in general. These supplemental material (Chapters 9 to 14) includes a glossary,

an index, lists of abbreviations and equations used in the document, a list of additional standards

and other documents related to DVI design, and a complete reference list of articles and reports

specific to each chapter used to develop the design guidance. These chapters are listed below.

• Chapter 2. How to Use this Document

• Chapter 3. General DVI Guidance for Level 2 and Level 3 Automation

• Chapter 4. Message Characteristics

• Chapter 5. Visual Interfaces

• Chapter 6. Auditory Interfaces

• Chapter 7. Haptic Interfaces

• Chapter 8. Driver Inputs

• Chapter 9. Glossary

• Chapter 10. Index

• Chapter 11. Abbreviations

• Chapter 12. Equations

• Chapter 13. Relevant Documents From the United States Department of Transportation,

SAE International, and International Organization for Standardization

• Chapter 14. References

Chapter 2. How to Use This Document

Two-Section Format

In this document, a consistent two-section format is used to present the individual human factors

topics provided in Chapters 3 to 8. On each page the chapter title is indicated by centered, bold

type within the header. As described in more detail below, the first section presents the title of the

topic; an introduction and overview of the topic; a high-level design goal; supporting design

guidance; a graphic, table, or figure that augments the text information; and the rating associated

with the topic. The second section provides the more detailed supporting rationale for the topic, as

well as special design considerations, cross-references to related topics, and a list of references. A

sample topic, with key features highlighted, is shown in Figure 2-1; a detailed description of the

presentation format of the topics follows.

Figure 2-1. Topic format used in the guidance document.

The First Section

The topic title is indicated by centered, bold type at the top of the first page of the section.

Introduction

This subsection briefly defines the topic and provides an overview of or background for the topic

area.

Design Goal

This subsection provides the high-level functional driver-vehicle interface implementation

objective for the topic. This design goal: (1) specifies an objective regarding driver responses or

activities that the driver-vehicle interface design is intended to support, and (2) includes the

primary automation application addressed by the topic (L2, L3 or both). The objective of this

section is to provide a goal without indicating the specific ways in which the design goal must be

met. Since there may be several design approaches that could achieve the functional outcomes

specified by the design goal, this level of guidance provides system and application developers

with flexibility for meeting the goal with alternative design and implementation approaches.

Design Guidance

This subsection provides the best-available design information from the literature, including

specific, quantitative design parameter values, if available, that can be incorporated into a driver-

vehicle interface that satisfies the design goal. This represents the most directly “actionable”

information presented in each topic, although the level of specificity may vary depending on

available research. A key goal within this subsection is to present the design assistance clearly

and succinctly, with a minimal amount of clutter. Where individual information in this subsection

reflects a direct quote, or has a direct source, the source is cited. Often, information presented

here reflects a synthesis of the findings, conclusions, or results from several sources, not just a

single source. Also, it may reflect the judgement of the authors, after the reviews and analyses of

the relevant data sources have been completed. In general, the Discussion subsection (discussed

below) is intended to provide users of this document with support and rationale for the design

information provided.

Figure, Table, or Graphic

This subsection provides a figure, table, or graphic to augment the design topic. This figure, table,

or graphic might take many forms, including: a drawing depicting a generic application of a

design principle or a design issue, a flowchart of measurement procedures for the design topic, a

table that summarizes the design topic, or schematic examples of particular visual warnings. The

figure, table, or graphic will provide “at-a-glance” information to support the use of the design

information.

The Second Section

Discussion

This subsection, which always starts on a new page, briefly summarizes the rationale behind the

choice of the supporting design guidance provided. The discussion can take many forms,

including a brief review of applicable empirical studies, references to traditional design practice,

or an analysis of relevant information. The discussion is presented primarily to help designers

understand the design guidance and to help them explain or justify the guidance to others

involved in developing a system or application.

Design Issues

This subsection presents special design considerations, design cases (e.g., older driver

capabilities), or other concerns that may impact the effectiveness of the driver-vehicle interface

design. Design issues are only included on an “as-available,” “as-needed” basis; not all topics

include a design issue subsection.

Cross References

This subsection lists the titles and page numbers of other topics within the guidance document

that are particularly relevant to the current topic.

Topic References

This subsection lists the references associated with the formulation of the design topic. Each of

these references will already have been noted within the text of the design topic, and assigned a

reference number. It provides a quick way for designers to identify the source of the design

information and for the authors to source the information.

Selection of Font Sizes in This Document

Note that the font sizes occasionally change in this document—this is deliberate. In general, the

individual DVI design topics in Chapters 3 to 11 are presented using a 10-point font to maximize

the content of the topics in as few pages as possible. In these chapters, the topic references might

be presented in an even smaller font, depending on space constraints. The remaining chapters,

including the tutorials and the “back matter” (i.e., glossary, index, references) are presented using

a 12-point font.

Chapter 3. General Design Guidance for Level 2 and Level 3 Automation

This chapter provides design guidance for facilitating the relationship between the driver and

automation systems so that drivers can engage the system appropriately. Drivers may benefit from

a good understanding of their automation system, including the signals and mechanisms for

transfer of control to the system and back again to the driver. This guidance addresses design

issues that support drivers in relinquishing primary control of their vehicle, assist them in retaking

the primary role of driving as workload changes, and monitoring the status of their system

activities. Of importance is the ability of the system to reconnect the driver when he or she has

become distracted by either secondary tasks or through the lack of focus on driving tasks that

follows handover of control to the system.

The following topics are included in this chapter.

• Current Automation Mode and Status

• Suitable Display Properties for Automation Mode and Status Messages

• Communicating Transfer of Control from Driver to System

• Communicating Transfer of Control from System to Driver

• Developing and Maintaining Driver Mental Models

• Automation Etiquette

• Special Considerations for Level 2 and Level 3 Automation

• Developing Driver Training Material for Automated Driving System Applications

Current Automation Mode and Status

Introduction

This topic provides information about communicating the current mode and status of the automation to drivers.

Automation mode refers to the type or level of automation that is active at a particular time. This includes the specific

driving functions that are automated and that are relevant for driver understanding of the system operation. The status

of automation refers to the information about the system overall, including mode, that is communicated to the driver.

Appropriate status feedback about automation mode and status is important for: (1) maintaining drivers’ situation

awareness, (2) communicating if their requests have been received by the automation, (3) informing drivers if the

system actions are being performed properly, or (4) informing drivers if problems are occurring (Toffetti et al., 2009).

Providing status information is a key part of meeting NHTSA’s human factors objectives of allowing drivers to safely

transition between automated and manual vehicle operation and communicating relevant information about the safe

operation of the vehicle effectively to the driver (Strickland, 2013).

Design Goal (L2, L3): Display the information that drivers need to maintain an understanding of the current

and impending automation state and modes.

Supporting Design Guidance

Based on the best available literature, the information in the table below shows the types of status information that

can be provided to the driver about the automation, and considerations for presenting this information.

Information Type What Information to Provide Why Information is Provided

System activation or

on/off status

A display indicating which automation mode

is currently active, if at all.

To support driver awareness of current

automation mode when the driver seeks

this information.

Mode transition

status

A display indicating that a transition in

automation mode is occurring or that one

will occur in the near future.

Under normal operating conditions, this

information is presented to help drivers

maintain awareness of the driving tasks.

Confirmation of

successful transfer

from automated to

manual control

A display or message confirming for the

driver that control has been transferred to the

driver as they would expect, or

communication of a failed/incomplete

transfer of control if the transfer is

unsuccessful.

To indicate a successful transfer of control

from the automation system to the driver.

System fault or

failure

A display or message indicating that part of

the system has failed or is not functioning

correctly.

To alert drivers that they must intervene in

and reclaim control of driving tasks that

have previously been performed by the

automation as a result of a system fault or

failure.

Discussion

A key consideration when developing automated driving systems is how to keep the driver aware of the current

driving situation when the automation is in control of a portion of the driving tasks. Providing mode and status

information about the automation can help drivers remain aware of the driving task even if he/she is not performing it

and can help drivers understand how the automation system works (Sarter, Woods, Billings, 1997).

System activation or on/off status: System activation or on/off information should be provided to the driver, even

though the driver may know this type of information about the system. In a simulator study investigating vehicles that

were going to be driven manually and automatically, Toffetti et al. (2009) provided drivers with a pre-activation

message, communicating that the eLane (automation) system would be active in 200m. Once the vehicle was in the

eLane area, a second message was provided to drivers indicating that the system was ready to use. Knowing the

automation system was active helped drivers know when they could transfer control to the automation since there

were multiple activation parameters that needed to be satisfied before the transfer could occur.

Mode transition status and confirmation of successful transfer from automated to manual control: If the automation

system detects a situation that it cannot handle, an impending transition in automation mode from system to driver

should be communicated if possible (in the case of L2 automation, and must be communicated in the case of L3

automation). This information should direct the driver’s attention to the transition. In the Toffetti et al. study (2009),

the system gave drivers a message a few minutes before deactivating. The majority of drivers (85%) reclaimed

control of the vehicle in response to the deactivation message while the other 15% of drivers did not. With L3

automation, drivers will need sufficient time to reclaim control of the vehicle as they may be engaged in non-driving

activities. Messages indicating the driver will need to take control soon can benefit drivers (Blanco et al., 2015; see

Chapter 4). After the transfer of control has occurred, the driver must receive information confirming that control has

been transferred back to him/her as expected (see topics 3-6 and 3-8). Mode transition status information should be

communicated to the driver under normal operating conditions to help the driver maintain awareness of system

function and driving tasks.

System fault or failure: When an automation function or system detects that it has stopped functioning as it was

designed to, this may need to be communicated to the driver depending on the effect upon the driver’s tasks. If only

part of the automated system stops functioning, resulting in partial automation, this may need to be communicated to

the driver as well. In response to a system failure, either partial or full, information regarding the specific failure

could be communicated if there are specific actions the driver needs to take in response to the failure. A study of L3

automation did not identify differences in driver responses to immediate take control requests based on the presence

or absence of a visible external hazard (Blanco et al., 2015; see Chapter 4).

Design Issues

Tracking or external input status: For automation systems that use external inputs, information could be

communicated to the driver indicating that the automation is detecting and receiving these necessary inputs. If the

system is not receiving these inputs, but the sensor or device that gathers these inputs is functioning properly, it may

indicate that something in the environment is preventing the sensor from gathering this information. An example of

this is a lane keeping system having trouble keeping the vehicle in the lane due to faded lane markings on the road. If

this information is provided to the driver, it should be provided in a way that does not interfere with the driving task

(Seppelt & Lee, 2007). The decision of whether to provide this information or not is dependent upon a number of

factors. Further research may assist with this decision by determining if drivers find the information useful and what

benefits and dis-benefits may be associated with its use.

Cross References

Suitable Display Properties for Automation Mode and Status Messages, 3-4; Communicating Transfer of Control

from Driver to System, 3-6; Communicating Transfer of Control from System to Driver, 3-8

Topic References

1. Norman, D. A. (1990). The problem of automation: Inappropriate feedback and interaction, not over-automation. Philosophical Transactions

of the Royal Society of London, B, 1-18.

2. Strickland, D. (2013, August 13). Every co-pilot needs a pilot (Speech for the Washington Automotive Press Association Luncheon).

Washington, DC. National Highway Traffic Safety Administration. Available at www.nhtsa.gov/sites/nhtsa.dot.gov/files/ds_wapa_8-14-

2013.pdf

3. Sarter, N. B., Woods, D. D., & Billings, C. E. (1997). Automation surprises. In G. Salvendy (Ed.), Handbook of Human Factors and

Ergonomics, Second Edition (pp. 1926-1943). Hoboken: Wiley & Sons.

4. Toffetti, A. Wilschut, E. S., Martens, M. H., Schieben, A., Rambaldini, A., Merat, N., & Flemisch, F. (2009). CityMobil: Human factor issues

regarding highly automated vehicles on eLane. Transportation Research Record: Journal of the Transportation Research Board, 2110, 1-8.

doi: 10.3141/2110-01

5. Blanco, M., Atwood, J., Vasquez, H. M., Trimble, T. E., Fitchett, V. L., Radlbeck, J., ... Morgan, J. F. (2015, August). Human factors

evaluation of level 2 and level 3 automated driving concepts (Report No. DOT HS 812 182). Washington, DC: National Highway Traffic

Safety Administration. Available at www.nhtsa.gov/sites/nhtsa.dot.gov/files/812182_humanfactorseval-l2l3-automdrivingconcepts.pdf

6. Seppelt, B. D., & Lee, J. D. (2007). Making adaptive cruise control (ACC) limits visible. International Journal of Human-Computer Studies,

65, 192-205.

Suitable Display Properties for Automation Mode and Status Messages

Introduction

This topic discusses message display properties for communicating automation mode and status information to the

driver. Status messages can be used in a range of ways and can have different functional objectives. Display

properties are important because the method used to communicate automation mode and status can be matched with

the nature of the information to facilitate communication and appropriate driver responses.

Design Goal (L2, L3): Present status information in a way that is compatible with the purpose of the

information being communicated and the expected driver response.

Supporting Design Guidance

The best available research suggests that this goal can be met when designers provide:

• Continuously available information on a dedicated display to communicate static status information that the

driver can access in a self-paced manner. Visual displays are suitable for this type of information because they

can be displayed continuously (Deatherage, 1972;

Seppelt, & Lee, 2007).

• Temporary, as needed, information with a clear onset in time to communicate time-dependent information,

especially a status that requires a driver to take action in the driving task. Auditory and haptic signals are

suitable for this type of information because these signals are omnidirectional and have a clear onset point

(Deatherage, 1972).

• Salient information to communicate important, time-critical, information that commands the driver’s

attention. Auditory and haptic signals are suitable for this type of information because they can quickly re-

orient the driver’s attention (Deatherage, 1972; Parasuraman & Riley, 1997).

• Non-salient information to communicate non-urgent, non-time-critical, information that notifies the driver

of automation mode and status, but does not interfere with the driving task. Visual displays are suitable for

this type of information because they can be viewed as the driving task allows and do not take the driver’s

attention away from the driving task (Deatherage, 1972; Arroyo, Sullivan, & Selker, 2006).

• A simple display to communicate basic information about the mode or status of the automation to the driver.

Auditory signals or visual displays are suitable for communicating this type of information, but choosing which

modality to use depends on the purpose of status information (Deatherage, 1972).

• A complex display to communicate detailed information about the mode or status of the automation to the

driver. Speech (for verbal messages) or visual displays are suitable for communicating this type of

information (Deatherage, 1972; Stanton, Dunoyer, & Leatherland, 2011).

Application of design guidance to specific status message types.

Type of Mode or

Status Information

Saliency Timing Complexity Example

System Activation or

On/Off Status

Non-salient

Continuously

available

Simple

A light indicating that the

automation system is on.

Mode Transition

Required

Salient

Temporary,

as needed

Simple

A tone and a “Place hands

on steering wheel” speech

message.

Confirmation of

Successful Manual

Override/Mode

Transition Success

Non-salient

Temporary,

as needed

Simple

An icon and a tone

indicating that the

automation will disengage

in the near future,

requiring the driver to

reclaim control of the

vehicle.

Tracking or External

Input Status

Non-salient

Continuously

available

Complex

A “radar” type display

indicating the vehicle that

the ACC system is

tracking and functioning

(Stanton, Dunoyer, &

Leatherland, 2011).

An Error or Bad Data

Resulting in Partial or

Imperfect Automation

Salient

Temporary,

as needed

Complex

A tone and a flashing icon

indicating to the driver that

the system is not

functioning normally and

that he/she needs to

reclaim control of the

vehicle.

Discussion

While it is important to communicate the current automation mode and status to drivers, the display properties used

to communicate this information are equally important. The following discusses the various display properties and

how they should be used to communicate mode and status information.

Providing continuously available information can help drivers understand how the automation system functions under

both normal and non-normal operating conditions, such as a failure, which is an important part of helping drivers

develop appropriate reliance on the automation system. If drivers understand how the system functions and they rely

on it, they are more likely to use it than if they do not understand how it functions and therefore do not rely on it

(Parasuraman and Riley, 1997). Seppelt and Lee (2007) investigated the use of an ecological interface design (EID)

display to provide a visual representation of ACC behavior. With the EID display, drivers were better able to respond

to ACC failures and the continuous information about the automation system and the roadway state helped drivers

rely on the automation appropriately when detecting and responding to the failures.

Continuously available information should be non-salient, meaning that it does not take the driver’s attention away

from the driving task, but should be available when the driver looks for it or have reached a point in the driving task

that they can take-in the information (Sarter, Woods, & Billings, 1997). The use of continuously available

information mostly applies to L2 automation, as with this level of automation, the driver is still expected to monitor

the roadway and for safe operation of the vehicle, meaning he/she will still be engaged in the driving task to a certain

extent.

Mode and status information that is presented in response to an event occurring, such as a take-over request from the

automation to the driver or a system failure, is different from continuously available information because it is often

time-dependent information and requires the driver to take action. This type of information should direct or reorient

the driver’s attention to the driving task. The communication of this information differs between Levels 2 and 3 of

automation. With L2, drivers are expected to be able to reclaim control from the automation on short notice, while in

L3 drivers are still expected to be able to reclaim control of the vehicle, yet on a more infrequent basis and with a

sufficient and more comfortable transition time than with L2. Toffetti et al. (2009) investigated the use an acoustic

interface and a vocal interface to indicate that the driver must reclaim control of the vehicle across three messages

(pre-message, first, and final message). In response to the automation deactivation message, at the pre-message step

more drivers took control in response to the vocal interface (55%) compared to the acoustic interface (40%) and after

the first message, more drivers with the acoustic interface (50%) took control compared to the vocal interface (35%).

Display complexity is another factor that must be considered when communicating automation mode and status

information. As mentioned on the previous page, a simple display should be used to communicate basic mode or

status information, while a complex display should be used to communicate more detailed mode or status

information. Stanton, Dunoyer, and Leatherland (2011) investigated the use of three interfaces to communicate that

the lead vehicle being tracked by the Stop & Go ACC system had changed. When using the “radar” type display,

drivers were better able to a detect change in the target vehicle being tracked by the Stop & Go ACC system than

when using the icon interfaces.

Design Issues

Beyond simple, salient, messages, sometimes messages must communicate a higher sense of urgency, e.g., a transfer

of control. In L3 automation, the message should be intrusive, but in a non-urgent way because the transfer of control

will happen in the future when driver have a lot of time to prepare. Conversely, in L2 automation drivers must take

control as soon as possible. Therefore, the message needs to be salient, attention-getting, and communicate a sense of

immediacy.

Cross References

Current Automation Mode and Status, 3-2; Selection of Sensory Modality, 4-6; Perceived Urgency of Auditory

Warnings, 6-2

Topic References

1. Deatherage, B. H. (1972). Auditory and other sensory forms of information presentation. In H. P. Van Cott & R. G. Kinkade (Eds.), Human engineering

guide to equipment design (rev. ed.). Washington DC: U.S. Government Printing Office.

2. Seppelt, B. D., & Lee, J. D. (2007). Making adaptive cruise control (ACC) limits visible. International Journal of Human-Computer Studies, 65, 192-

205.

3. Parasuraman, R., & Riley, V. (1997). Humans and automation: Use, misuse, disuse, abuse. Human Factors, 39(2), 230-253.

4. Arroyo, E., Sullivan, S., & Selker, T. (2006). CarCoach: A polite and effective driving coach. Proceedings of CHI ’06 Extended Abstracts on Human

Factors in Computing Systems, 357-362.

5. Stanton, N. A., Dunoyer, A., & Leatherland, A. (2011). Detection of new in-path targets by drivers using stop & go adaptive cruise control. Applied

Ergonomic, 42, 592-601.

6. Sarter, N. B., Woods, D. D., & Billings, C. E. (1997). Automation surprises. In G. Salvendy (Ed.), Handbook of Human Factors and Ergonomics,

Second Edition (pp. 1926-1943). Hoboken: Wiley & Sons.

7. Toffetti, A., Wilschut, E. S., Martens, M. H., Schieben, A., Rambaldini, A., Merat, N., & Flemisch, F. (2009). CityMobil: Human factor issues

regarding highly automated vehicles on eLane. Transportation Research Record: Journal of the Transportation Research Board, 2110, 1-8. doi:

10.3141/2110-01

Communicating Transfer of Control From Driver to System

Introduction

This topic discusses the driver’s information needs within a driver-initiated transfer of control from the driver to an

automated system and what information supports the driver during this transition. This type of transfer reflects a

series of operations through which the driver transfers part of or the entire driving task to the automated system.

System information during these transfers is important to maintain a driver’s calibrated trust in the system and to help

drivers maintain awareness of driving tasks and the driving situation.

Design Goal (L2, L3): Support the transition from manual to automated driving by acknowledging a driver’s

request to engage automation and provide information about the status of the transfer of control.

Supporting Design Guidance

The best available research on this topic suggests that this design goal can be met when the following points are

considered:

The current system status is provided at all times (Toffetti et al., 2009).

Automation engagement requests are acknowledged upon receipt to prevent duplicate or conflicting

inputs from the driver, and to prevent the driver from releasing control without the automation being

active.

Feedback acknowledging a driver automation activation request is provided within 250 ms of input

(ISO 15005, 2002; Alliance of Automobile Manufacturers, 2006).

If the transfer was successful, provide the driver with a notification and update the displayed status of

the automation.

If the transfer was unsuccessful, the driver is provided with a notification as to the failure of the

automation to engage and why the automation did not engage (Tsao, Hall, & Shadlover, 1993; Merat

& Jamson, 2009).

The use of unimodal or multimodal notifications and messages is based on the context of the

situation.

Distinctive messages are used for successful and unsuccessful transfer of control events.

Display needs during driver to automation transfer of control.

This flowchart shows a hypothetical driver-automation interaction. This topic addresses what information the system

can provide to support the driver during the interaction, indicated within the shaded box. The automation status is

displayed at all times. Following an activation request, the automation displays the success of the transfer or, in the

case that automation does not activate, displays that automation did not activate and what prevented activation from

occurring. System to driver transfers are covered in Communicating Transfer of Control from System to Driver, 3-8.

Discussion

This topic addresses how the DVI can support the driver’s information needs during a manual-to-automated transition

once the driver has initiated the request for automated control. Although well examined in other domains, the

activation of automation and transfer of control from the driver to the automated driving system have not been

extensively examined for L2 or L3 systems. However, the basic guidance of human factors, as well as the extant

research from L1 and exploratory L2 research, can inform the DVI designer as to some basic guidance for supporting

the driver to automation transfer of control.

Providing the current status at all times can help prevent mode awareness failures (Toffetti et al., 2009). The

automation should provide an acknowledgement of the request in a timely fashion if the request should be expected.

Providing rapid feedback to the driver helps to avoid driver uncertainty. Delayed responses can lead to drivers

providing second inputs or becoming distracted by the systems’ non-response. Standards organizations have provided

some guidance for system response time maximal limits for drivers’ interaction with vehicle user interfaces that

provide good guidance for designers of automated driving system DVIs. ISO 15005 (2002) provides a limit of 250

ms for providing feedback to the driver. This value is also used by the Alliance of Automobile Manufacturers (AAM)

as a maximum system response time for in-vehicle information and communications systems providing a response to

drivers (AAM, 2006).

In addition to acknowledging a driver’s input, research from other areas of human-automation interaction suggest that

it is important to include information on the success or failure of the automation activation request (Sarter & Woods,

1995). This can reduce the likelihood of a mode awareness failure, where the driver believes the automation activated

and ceases control of the vehicle without automation having assumed control. If the transfer is unsuccessful even

though the system displayed that automation was available, drivers may benefit from receiving information about the

reason for the failure. Drivers may attempt to immediately reactivate the automation. If a reason for the failure is

something that can be provided in a simple manner and addressed by the driver, it may help the driver successfully

modify vehicle conditions (e.g., centering the vehicle in the lane, stopping at an appropriate location) necessary for

automation activation.

Design Issues

The use of a unimodal or multimodal notifications or messages should be considered based on the context in which

the message is being generated. For uneventful driver-initiated activations of automation, basic acknowledgments

may suffice. If a driver may misinterpret a scenario, such as a failed automation activation request, consider using a

multimodal display. Distinctive messages are needed for successful and unsuccessful transfer of control events to

minimize the potential for confusion as to the transfer status. The use of unique coding can help create distinctive

messages for these two events.

Cross References

Current Automation Mode and Status, 3-2; Suitable Display Properties for Automation Mode and Status Messages, 3-4;

Communicating Transfer of Control from System to Driver, 3-8; Selection of Sensory Modality,4-6; Multimodal

Messages, 4-8

Topic References

1. Toffetti, A. Wilschut, E. S., Martens, M. H., Schieben, A., Rambaldini, A., Merat, N., & Flemisch, F. (2009). CityMobil: Human factor issues

regarding highly automated vehicles on eLane. Transportation Research Record: Journal of the Transportation Research Board, 2110, 1-8.

doi: 10.3141/2110-01

2. ISO 15005. (2002). Road vehicles—Ergonomic aspects of transport information and control systems—Dialogue management principles and

compliance procedures. Geneva: International Organization for Standards.

3. Alliance of Automobile Manufacturers. (2006). Statement of principles, criteria and verification procedures on driver interactions with

advanced in-vehicle information and communication systems, including 2006 updated sections (Report of the Driver Focus-Telematics

Working Group). Retrieved from www.autoalliance.org/index.cfm?objectid=D6819130-B985-11E1-9E4C000C296BA163.

4. Tsao, H.-S. J., Hall, R. W., & Shadlover, S. E. (1993). Design options for operating fully automated highway systems. Berkeley: University of

California PATH Institute of Transportation Studies.

5. Merat, N., & Jamson, A. H. (2009). How do drivers behave in a highly automated car? Proceedings of the 5th International Driving

Symposium on Human Factors in Driver Assessment, Training, and Vehicle Design, Big Sky, MT, June 22-25, 2009 514-521.

6. Sarter, N. B., & Woods, D. D. (1995). How in the world did we ever get into that mode? Mode error and awareness in supervisory control.

Human Factors, 37, 5-19. doi: 10.1518/001872095779049516

Communicating Transfer of Control from System to Driver

Introduction

This topic discusses the driver’s information needs of a system-initiated transfer of control from an automated system

to the driver and what information supports the driver during this transition. Transfer of control can range from

gradual and expected by the driver to immediate and unexpected. Consequently, providing drivers with information

to support a safe transition from automated to manual vehicle control in all circumstances may be beneficial.

Design Goal (L2, L3): Support the transition from automated to manual driving by providing information

about the need for the driver to take manual control.

Supporting Design Guidance

The best available research on this topic suggests that this design goal can be met when the following

points are considered:

• The driver is provided with information on when they need to take control (Gold, Damböck,

Lorenz, & Bengler, 2013; Blanco et al., 2015).

• The driver is provided with information on how to take control if a specific control input is

required (Toffetti et al., 2009).

• The driver is provided with information on why the driver needs to take control. For time-

critical situations, this may be a simplified “take control” message. For less time-critical

situations, more information may be provided.

• The current system status is provided at all times, allowing the driver to validate the

disengagement (Sheridan & Parasuraman, 2005).

• Notifications and messages related to a “take control” message are multimodal. (Blanco et

al., 2015; Toffetti et al., 2009; Brookhuis, van Driel, Hof, van Arem, & Hoedemaeker,

2008).

•

Any collision avoidance system messages are provided to the driver, if applicable.

Display needs during automation to driver transfer of control.

This flowchart shows a hypothetical driver-automation interaction. This topic addresses information the system can

provide to support the driver during the interaction, indicated within the shaded box. The urgency of the request is

considered in what type of message to provide to the driver. The reason for the take control message is provided

depending on the time critical nature of the request. Driver to system transfers are covered in Communicating

Transfer of Control from Driver to System (3-6).

Discussion

The “take control” request should inform the driver about when to take control. A CityMobil study examining an L2-

type automated driving included a three-stage warning prior to emergency braking at the end of the automated travel

lane (Toffetti et al., 2009). Many drivers took manual control at the first warning issuance, and the majority of drivers

had taken control after the second warning. Providing earlier “take control” requests can aid drivers in transitioning to

manual control. A simulator evaluation of L3 driving used a “take control” scenario with either a 5 s or 7 s time-to-

collision (TTC) used as the “take control” request issuance point (Gold, Damböck, Lorenz, & Bengler, 2013. The

amount of steering maneuvers increased with a greater amount of warning (i.e., the 7 s. TTC) and the amount of

braking maneuvers decreased. Drivers in the 7 s. TTC condition could sample more roadway information and thus

steer the vehicle around the obstacle. In an L3 study where a 40 s duration staged alert was provided, drivers took

control on average 17.0 s post-alert issuance; all drivers took control before the alert escalated to a more urgent level

(Blanco et al., 2015; see Chapter 4).

Information about how the driver should take control should be provided, if applicable. If the driver should perform a

certain input, such as taking manual control of steering (Toffetti et al., 2009) this information should be clearly

provided. Guidelines from the Automated Highway System project recommended against limiting the methods a

driver can use to take manual control of the vehicle (Levitan, Bunus, Dewing, Reinhart, Vora, & Llaneras, 1997), and

cruise control (an L0 automation system most drivers are familiar with) similarly allows drivers to take manual

control through multiple means.

The status of L2 or L3 automation following a temporary driver manual input should be clearly communicated to the

driver (indicating if the automation is still active and in control) (Sheridan & Parasuraman, 2005). As a safety critical

message, take control messages should be provided in a multimodal manner. Simulator studies of L1-type automated

systems have used a visual icon for system status accompanied by an auditory signal to indicate when the system

status changed (Brookhuis, van Driel, Hof, van Arem, & Hoedemaeker, 2008). Evaluations of L2-type automation

have used a constant visual status message with visual and auditory notifications for transition states (Toffetti et al.,

2009), or a visual-tactile combination (Blanco et al., 2015; see Chapter 2) to indicate this status.

Design Issues

When communicating transition information to the driver, a key question is how to balance the complexity of

information provided against the time available to provide this information. In the absence of research specific to L2

and L3 automation, guidance from driver warnings (see topic 4-4) should be followed and information should be

presented in the simplest manner supporting a timely and effective driver response. Providing the reason for the “take

control” request may be beneficial for supporting mental models. While potentially inappropriate during urgent “take

control” requests, information about the reason for the “take control” request may prevent the driver from attempting

to reengage automation in the presence of a limiting condition and helps support the formation of functionally-

accurate mental models of the system (see topic 3-10). Providing the current automation status allows drivers to

verify the automation disengaged (Sheridan & Parasuraman, 2005). Relevant safety warnings (e.g., collision

warnings) should be provided.

The guidance in this topic does not address transfers between L2 and L3 automation. Insufficient research exists to

provide an understanding of what is needed to support transitions between higher automation levels (i.e., L3 to L2)

and how to ensure the driver begins performing the monitoring tasks required at that point. ACC provides a similar

parallel in that SAE standards do not permit ACC to transition to normal cruise control (SAE J2399, 2003). However,

some simulator research suggests that drivers may benefit from a supportive level of automation during a transition

(Gold, Lorenz, Damböck, & Bengler, 2013).

Cross References

Current Automation Mode and Status, 3-2; Suitable Display Properties for Automation Mode and Status Messages,

3-4; Communicating Transfer of Control from System to Driver, 3-8; Message Complexity, 4-4; Developing and

Maintaining Mental Models, 3-10

Topic References

1. Gold, C., Damböck, D., Lorenz, L., & Bengler, K. (2013). “Take over!” How long does it take to get the driver back into the loop? Proceedings of the

Human Factors and Ergonomics Society 57th Annual Meeting, 1938-1942. doi: 10.1177/1541931213571099

2. Blanco, M., Atwood, J., Vasquez, H. M., Trimble, T. E., Fitchett, V. L., Radlbeck, J., ... Morgan, J. F. (2015, August). Human factors

evaluation of level 2 and level 3 automated driving concepts (Report No. DOT HS 812 182). Washington, DC: National Highway Traffic

Safety Administration. Available at www.nhtsa.gov/sites/nhtsa.dot.gov/files/812182_humanfactorseval-l2l3-automdrivingconcepts.pdf

3. Toffetti, A., Wilschut, E. S., Martens, M. H., Schieben, A., Rambaldini, A., Merat, N., & Flemisch, F. (2009). Human factors issues regarding highly

automated vehicles on eLane. Transportation Research Record: Journal of the Transportation Research Board, 2110, 1-8. doi: 10.3141/2110-01

4. Sheridan, T. B. and Parasuraman, R. (2005). Human-automation interaction. Reviews of Human Factors and Ergonomics, 1, 89-129.

5. Brookhuis, K. A., van Driel, C. J. G., Hof, T., van Arem, B., & Hoedemaeker, M. (2008). Driving with a congestion assistant; mental

workload and acceptance. Applied Ergonomics, 40, 1019-1025. doi: 10.1016/j.apergo.2008.06.01

6. Levitan, L., Bunus, M., Dewing, W. L., Reinhart, W., Vora, P. B., & Llaneras, R. E. (1997). Preliminary human factors guidelines for automated

highway system designers (2nd Ed.) (Report No. FHWA-RD-97-125). McLean, VA: Federal Highway Administration.

7. SAE J2399. (2003). Adaptive cruise control (ACC) operating characteristics and user interface. Warrandale, PA: SAE International.

8. Gold, C., Lorenz, L., Damböck, D., & Bengler, K. (2013, November). Partially automated driving as a fallback level of high automation. Paper

presented at Tagung Fahrerassistenzsysteme: Der weg zum automatischen fahren. Retrieved from http://mediatum.ub.tum.de/doc/1187198/1187198.pdf

Developing and Maintaining Driver Mental Models

Introduction

This topic discusses considerations for designing an automation system that supports drivers developing and

maintaining a functionally accurate mental model of how the automation system operates. A mental model

encompasses the user’s knowledge of an automated system’s purpose, how it functions, and how it is likely to

function in the future (Goodrich & Boer, 2003). Drivers with a functionally accurate mental model of the automation

may be more likely to avoid errors based on incorrect assumptions about system operation, recognize a failure if it

occurs, and use the automation appropriately.

Design Goal (L2, L3): Support the development and maintenance of a functionally accurate mental model of

the automated system through design and training.

Supporting Design Guidance

The best available research on this topic suggests that this design goal can be met when the following points are

considered:

Provide training or information to support the formation and refinement of a functionally accurate

mental model (Redding, Cannon, & Seamster, 1992; Cassidy, 2009).

Provide information about the system at a level of detail that can be understood. The automation may

best be described as a set of tasks or functions, not in terms of sensors and control systems. The

information’s level of specificity may be geared towards relevant details for the DVI and the driver’s

direct experience with the system without the inclusion of unnecessary details. [4]

Automation works in a consistent manner and training material are consistent with the automation’s

operation.

Presenting information in a way that supports drivers’ mental models.

Suggested Not Suggested

Example for a Hypothetical

L2 Highway-speed System

Describe automation as a set of tasks

or functions and how they relate to

the driver and driving task.

Describe automation in terms of the

sensors and controllers.

Not Recommended: Training manual

describes the system as using a near-

infrared camera and machine vision

algorithms to detect lane markings.

Better: Training manual describes

the system as looking for lane

markings.

Present information at the most basic

level that provides the needed

information.

Present jargon or complex

information, such as technical

information, unnecessarily.

Not Recommended: Training manual

describes the ACC function as a

sensor fusion between machine

vision- and radar-based sensors.

Better: Training manual describes

the system as constantly measuring

the distance between your vehicle

and objects in front of your vehicle.

Use an appropriate level of

specificity, focusing on relevant

details.

Provide unnecessary details or

needless information.

Not Recommended: DVI provides

status and continuously updated

distance values for lane keeping and

headway separately.

Better: DVI provides a single

presentation of automation status.

Discussion

Drivers have mental models that help them understand a system’s current performance as well as extrapolate its

likely future behavior (Goodrich & Boer, 2003). While drivers have existing mental models for common automation

features, such as cruise control, they will have vague or non-existent mental models for the first implementations of

L2 and L3 automation. Thus, mental models will be developed through instruction, training, and use.

Unfamiliar automation can be paired with instruction and training to support the formation, refinement, and

maintenance of the driver’s mental model. Research in human-automation interaction, such as air traffic control, has

identified that having a functionally accurate mental model for automated systems is a central aspect of reaching a

level of expertise with the system (Redding, Cannon, & Seamster, 1992). Additionally, having a functionally accurate

mental model affects an operator’s level of trust in an automated system (Cassidy, 2009). Automated decision aid

research has suggested that operators with a functionally accurate mental model of an automated system are better

able to identify automation errors (Wilkison, 2008).

A simulator study of an L1 automation technology (ACC) examined the formation of drivers’ trust and conceptual

models over 10 days (Kazi, Stanton, Walker, & Young, 2007). Results indicated that drivers’ conceptual models

tended to be consolidated by the fourth day. Additionally, despite specifically stating that the ACC system did not

include a collision avoidance feature, 33 percent of participants incorrectly described collision avoidance as a system

component. Further analysis indicated that some drivers were confusing the headway maintenance feature of ACC

with a collision avoidance maneuver. While not representative of ACC’s real-world performance, this research

highlights the importance of information presentation.

In advanced systems, some complex information cannot be processed until it has been stored in a schematic form in

long-term memory (Pollock, Chandler, & Sweller, 2002). Depending on the implementation, automation may be a

highly complex system from a human-automation interaction perspective. Some best practices from prior human

factors research into automation are applicable, including presenting information at an appropriate level and at the

appropriate level of detail (Wiener, Chute, & Moses, 1999).

Support formation of functionally accurate mental models by having automation work in a consistent manner. In the

case of the L1 ACC automation technology, Goodrich and Boer (2003) note that the utility of the automation is a

function of not only the driver’s understanding of the vehicle-environment interaction but also the driver’s

understanding of the automation. When the ACC system behaves in a manner consistent with that of a reasonable

driver, the driver’s mental model of the system will be supported and the driver will be more likely to intervene when

needed (Goodrich, Boer, & Inoue, 1999). While this research examined the case of an L1 automated system, the same

guidance likely applies to other monitored automation types that fall under L2. Additionally, any information about

the system provided to the driver should be consistent with how the system will function.

Design Issues

A driver’s mental model of how a system operates, performs, and will likely perform helps determine the likelihood

that the system will be ignored, used correctly, or misused/misapplied. Drivers of the first L2 and L3 automated

driving systems will likely have weak, incorrect, or absent mental models of early automated systems. Further, it is

unknown what differences may manifest between drivers’ mental models of L2 and L3 automation. Given the myriad

ways drivers can interact with automation, design should be sensitive to, and supportive of, the formation of

functionally accurate mental models.

Cross References

Current Automation Mode and Status, 3-2; Developing Driver Training Material for Automated Driving System

Applications, 3-14

Topic References

1. Goodrich, M. A., & Boer, E. R. (2003). Model-based human-centered task automation: A case study in ACC system design. IEEE

Transactions on Systems, Man, and Cybernetics, Part A, Systems and Humans, 33(3), 325-336.

2. Redding, R. E., Cannon, J. R., & Seamster, T. L. (1992). Expertise in air traffic control (ATC): What is it, and how can we train for it?

Proceedings of the Human Factors and Ergonomics Society, 36, 1326-1330.

3. Cassidy, A. M. (2009). Mental models, trust, and reliance: Exploring the effect of human perceptions on automation use (Master’s thesis).

Naval Postgraduate School, Monterey, CA.

4. Pollock, E., Chandler, P., & Sweller, J. (2002). Assimilating complex information. Learning and Instruction, 12, 61-68.

5. Wilkison, B. D. (2008). Effects of mental model quality on collaborative system performance (Master’s thesis). Georgia Institute of

Technology, Atlanta, GA.

6. Kazi, T. A., Stanton, N. A., Walker, G. H., & Young, M. S. (2007). Designer driving: Drivers’ conceptual model and level of trust in adaptive

cruise control. International Journal of Vehicle Design, 45(3), 339-360.

7. Wiener, E. L., Chute, R. D., & Moses, J. H. (1999). Transition to glass: Pilot training for high-technology transport aircraft (NASA/CR-

1999-208784). Washington, DC: National Aeronautics and Space Administration.

8. Goodrich, M. A., Boer, E. R., & Inoue, H. (1999). A model of human brake initiation behavior with implications for ACC design. Proceedings

of IEE/IEEJ/JASI International Conference on Intelligent Transportation Systems, 86-91.

Special Considerations for Level 2 and Level 3 Automation

Introduction

This topic discusses special considerations associated with L2 and L3 automation. The level of automation can

determine how involved the driver must be in monitoring the environment to ensure the safe operation of the vehicle.

Humans, however, are poor at performing vigilance and monitoring tasks. Automation may benefit from being

designed to provide information to drivers in a way that promotes a level of supervision by the driver that is

appropriate for the level of automation.

Design Goal (L2, L3): Encourage adequate driver supervision relative to the requirements of the level of

automation.

Supporting Design Guidance

Potential challenges to drivers using L2 and L3 automation arise from driver task underload. The following guidance

is based on the best available research on this topic and provide general information related to addressing these

challenges.

Automation that optimizes drivers’ workload may promote situation awareness and assist in

maintenance of an appropriate level of vigilance during automated driving.

Automation that may need driver intervention may benefit from having an awareness of the driver’s

status as part of supporting transitions from automated to manual driving. There may be benefits from

employing strategies that engage the driver’s attention when deemed necessary.

L2 automation may perform best when designed to:

− Clearly indicate the level or mode of automation

− Encourage drivers to attend to forward roadway conditions (Blanco et al., 2015; see Chapter 3)

− Provide alerts for system failures and limitations (Salinger, 2012; Hoeger et al., 2011)

L3 automation may perform best when signals used to indicate or request a transition from

automation to manual control are salient, distinctive, and unambiguous in order to capture drivers’

attention and promote understanding of the automation state.

A. Display indicating request to

reclaim control from automation.

B. Message to capture

inattentive drivers’ attention.

C. Flashing LEDs in the windscreen

at right to capture inattentive

drivers’ attention.

A. Figure 263, use and modification of fifth image, B. and C. from Figure 133]. Adapted from Hoeger et al. (2011). Used with permission.

Discussion

L2 automation requires drivers to monitor the road and remain able to take control of the vehicle at any time. L2

automation takes drivers out of the vehicle-control loop and requires them to adopt a supervisory role; however,

humans generally perform poorly in the role of monitors (Kaber, & Endsley, 2004). Out-of-the-loop (OOTL)

problems include complacency, vigilance decrements, loss of situation awareness, and decay of skill when relying on

high-level automation compared with low- or intermediate-level automation (Kaber, & Endsley, 2004; Stanton,

Young, Walker, Turner, & Randle, 2001). Stanton et al. (2001) found that drivers were less able to reclaim control of

the vehicle from the automation in an emergency braking scenario when both ACC and a lane centering system

(LCS) were active. Similarly, drivers’ brake reaction times in a simulator study (Young & Stanton, 2007) were 1.0 s

to 1.5 s longer for OOTL drivers using ACC versus manual headway keeping. Strategies should be adopted that

promote situation awareness and support understanding of the automation state when drivers are out of the control

loop. Such strategies may include interactive, periodic, messages that serve to engage the driver (Widmann, Salinger,

Dufour, and Green, 2011), or messages based on driver monitoring systems (Blanco et al., see Chapter 3; Hoeger et

al., 2011).

L3 automation requires the automation to assume a greater role for the safe operation of the vehicle, and the driver is

not expected to constantly monitor the road. Consequently, the requirements for vigilance and situation awareness are

reduced, but not eliminated, as compared with L2 automation. As drivers do not have to engage in driving, they may

become so involved in secondary tasks (e.g., eating, reading, watching entertainment, or perhaps even sleeping) that

they may entirely miss the transition message. In a driving simulator study, Carsten, Lai, Barnard, Jamson, and Merat

(2012) found that drivers in vehicles with combined ACC and LCS engaged in activities other than driving (e.g.,

eating, listening to the radio, watching DVDs, etc.) when the automation was active, and the level of inattention to the

driving environment increased proportionally with level of automation. Similarly, drivers in simulator and test track

experiments (Salinger, 2012) performed relatively risky tasks, such as cell phone talking, texting, and emailing;

watching DVDs; and reaching into the back compartment, during automated driving. Therefore, it may be of benefit

to understand the driver’s status before alerting them to a forthcoming transition. An important issue for L3

automation, then, is ensuring that the signal presented by the automation to inform drivers of the change of

automation is sufficiently salient, distinctive, and unambiguous that drivers are able to quickly and easily understand

the state of the automation.

Design Issues

Hoc, Young, and Blosseville (2009) assert that automation should be implemented as a human-automation

cooperative system rather than simply using a task-based framework. The goal of a cooperative system is to optimize

the driver’s task load in order to avoid driver overload or OOTL problems. The European HAVEit demonstration

project (Salinger, 2012) adopted a cooperative approach by integrating driver monitoring systems with incremental

levels of automation (L0 through L2). When driver state detection algorithms determined that drivers were drowsy or

inattentive, the system presented visual and/or auditory signals to capture drivers’ attention and bring them back into

the vehicle control or monitoring loop. Another approach to driver monitoring is to periodically provide a message to

which the driver must respond. Widmann, Salinger, Dufour, and Green (2011) used elbow switches, foot switches,

and steering wheel touch sensing as inputs to allow the driver to respond to these messages.

Cross References

Communicating Transfer of Control from Driver to System, 3-6; Communicating Transfer of Control from System to

Driver, 3-8; Message Complexity, 4-4; Developing and Maintaining Mental Models, 3-10; Chapter 5: Visual

Interfaces; Chapter 6: Auditory Interfaces; Chapter 7: Haptic Interfaces

Topic References

1. Blanco, M., Atwood, J., Vasquez, H. M., Trimble, T. E., Fitchett, V. L., Radlbeck, J., ... Morgan, J. F. (2015, August). Human factors

evaluation of level 2 and level 3 automated driving concepts (Report No. DOT HS 812 182). Washington, DC: National Highway Traffic

Safety Administration. Available at www.nhtsa.gov/sites/nhtsa.dot.gov/files/812182_humanfactorseval-l2l3-automdrivingconcepts.pdf

2. Salinger, J. (2012). Human factors for limited-ability autonomous driving systems. Retrieved from

http://onlinepubs.trb.org/onlinepubs/conferences/2012/Automation/presentations/Salinger.pdf

3. Hoeger, R., Zeng, H., Hoess, F., Kranz, T., Boverie, S., Straus, M., Jakobsson, E., … & Nilsson, A. (2011). The future of driving. Deliverable D61.1.

Final report. Version 1.0 (Report for the Highly Automated Vehicles for Intelligent Transport [HAVEit] 7th Framework Programme). Retrieved from

www.haveit-eu.org/displayITM1.asp?ITMID=24&LANG=EN

4. Kaber, D. B., & Endsley, M. R. (2004). The effects of level of automation and adaptive automation on human performance, situation awareness and

workload in a dynamic control task. Theoretical Issues in Ergonomics Science, 5(2), 113-153. doi: 10.1080/1463922021000054335

5. Stanton, N. A., Young, M. S., Walker, G. H., Turner, H., & Randle, S. (2001). Automating the driver’s control tasks. International Journal of Cognitive

Ergonomics, 5(3), 221-236.

6. Young, M. S., & Stanton, N. A. (2007). Back to the future: Brake reaction times for manual and automated vehicles. Ergonomics, 50(1), 46-58. doi:

10.1080/00140130600980789

7. Widmann, G., Salinger, J., Dufour, R., & Green, C. (2011). Human factors studies for limited-ability autonomous driving systems (LAADS).

Presentation at the SAE 2011 Government/Industry Meeting. Washington DC. Retrieved from www.sae.org/events/gim/presentations/2011.

8. Carsten, O., Lai, F. C. H., Barnard, Y., Jamson, A. H., & Merat, N. (2012). Control task substitution in semiautomated driving: Does it matter what

aspects are automated? Human Factors: The Journal of the Human Factors and Ergonomics Society, 54(5), 747-761. doi: 10.1177/0018720812460246

9. Hoc, J.-M., Young, M. S., & Blosseville, J.-M. (2009). Cooperation between drivers and automation: Implications for safety. Theoretical Issues in

Ergonomics Science, 10(2), 135-160.

Developing Driver Training Material for Automated Driving System Applications

Introduction

Training is the process by which we acquire knowledge and skill on specific topics, systems, or applications. Many

driver training programs have been developed by using curricula designers that “believe should be there” and have

been minimally influenced by contemporary findings from the behavioral sciences (Brock, McFann, Inderbitzen, &

Bergoffen, 2007). This topic summarizes recent findings from the behavioral research literature on several training

methods that could be used to train drivers on L2 and L3 automated driving systems. This DVI design topic may be

most applicable to automation applications that are especially novel or complex.

Design Goal (L2, L3): Systematically design, develop, and evaluate training for automated systems in a

manner consistent with the goals of the system.

Supporting Design Guidance

The best available research on this topic suggests that this design goal can be met when the following points are

considered:

Provide information on system functionality and system limits in a driver training program (Panou,

Bekiaris, & Touliou, 2010).

For complex systems with many features, variable priority training (VPT) may be used. In VPT,

trainees are presented a complex task divided into subtasks between which they vary their

attentional priority (Boot et al., 2010; Wickens, Hutchins, Carolan, & Cumming, 2013) (e.g.,

divide the task into Parts A and B. Start with the trainee devoting 80% of their attention to Part A

and 20% to Part B. Then, dedicate 20% to Part A and 80% to Part B. Finally, practice with 50%

attention to both.).

Use error-training as a strategy to reduce driver overconfidence in using an automated system and

to influence trainees to generate their own coping strategies for novel situations not covered in

training (Ivancic, & Hesketh, 2000).

Training on the use and function of in-vehicle systems occurs after novice drivers have obtained

the basic and rudimentary skills needed for safe-driving (Panou, Bekiaris, & Touliou, 2010).

Simple 5-stage iterative training design model (Graving, Easterlund, & Manser, 2011).

Content: During the initial design phases of developing a training curriculum, a first step is to obtain

training content and information about general training requirements. This can be obtained through

user interviews, use-case scenarios, focus groups, and task analyses. Literature searches can be useful

for preparing for user-testing and obtaining information about driver tasks (e.g., previous task

analyses, McKnight, & Adams, 1970). The other phases of the iterative design process can also be

used to generate content for training. Prototype: Start by generating simple prototypes (e.g., paper-

based drafts of a training manual) during this phase to test training concepts and ensure the training

Content

Prototype

Hand-off

Training

Test

Cycle

• Literature Review

• Focus group

• One-to-ones

• Task analysis

• Testing

• Training Manual

• Simulator Scenarios

• Focus group

• Pilot test

• Train-the-trainers

• Observe training

content is appropriate. During subsequent iterations, generate more complex prototypes (e.g., pilot-

test class-room instruction or develop simulator-based scenarios) to test the functional aspects of the

training program—e.g., timetables for when certain topics are covered, methods of presentation.

Test: The process of testing a prototype can be useful for obtaining additional design requirements as

well as uncovering implementation issues. Testing can include observations of trainers carrying out

the training program.

Next Cycle: A decision phase where the testing results are used to decide if additional iterations are

required.

Hand-off Training: If testing indicates that iterations are no longer required, the training program is

ready to be handed off to trainers.

Discussion

While simpler automated systems may not require specific training, some more complex automated systems may

require initial training at vehicle delivery to ensure the driver has a minimum level of operating skills with the

automation prior to use. Training system development should be systematic, deliberate, and informed by research

findings from the instructional system design literature.

A training program may cover all features of the automation (especially if only a single type of automation is present

in the vehicle) or can entail different aspects of the system itself. Training sessions should focus on the functionality

of the system, including information on limitations (including when automation may not be activated or used) and

any potential for malfunction (Panou, Bekiaris, & Touliou, 2010). Such training programs could highlight any

significant limitation of the system for informing the driver of sudden hazards, as well as limitations based on sensing

technologies used. For instance, drivers should be made aware that the highway speed automation will not change

lanes to pass slower moving lead vehicles, or that the traffic jam assist feature will not remain active after a certain

continuous period of time above a set speed.

Training programs should also include information on how to read and interpret the DVI, methods for activating the

system, available action options for manual overrides and input during automated driving (e.g., accelerate, brake,

steer, or a combination of inputs), how the system reacts to manual inputs during automated driving, and what types

of control transfers may be encountered. Proper training should also provide drivers with the ability to transform and

incorporate declarative knowledge (e.g., knowledge about the system) into more automatic-like actions for using the

automation. Such behavioral skills can become relatively immune from deterioration in a broad range of contexts

(Rasmussen, 1983).

Different training strategies may be employed to ensure that drivers gain knowledge on and retain useful driving

strategies. The VPT strategy is an effective training strategy for complex tasks (Graving, Easterlund, & Manser,

2011). The method presents trainees with the whole task, which is maintained during training, but different

components are systematically emphasized or deemphasized to allow more attention to be focused on specific parts

while still preserving the necessary element of time-sharing of attention across the whole task (Wickens, Hutchins,

Carolan, & Cumming, 2013). Drivers trained with this strategy tend to learn faster and reach higher levels of mastery

compared with training programs that emphasize all components equally (Boot et al., 2010; Wickens, Hutchins,

Carolan, & Cumming, 2013). In addition, the error-training strategy where learners acquire information about a task

through exploration, testing self-generated hypotheses, and trial-and-error may be effective in reducing

overconfidence and can help drivers develop coping strategies for scenarios not covered in training (Ivancic, &

Hesketh, 2000).

In some cases, different training settings may be necessary. Drivers 70 to 89 years old benefit more after training is

complete when instructors provide feedback on actual driving in addition to classroom training (Porter, 2013). As

much as possible, training should provide drivers with an ability to manage the traffic events and scenarios presented

during training as well as novel events not encountered during training.

Design Issues

Large in-vehicle displays with additional computing capabilities are becoming more common, providing a potential

platform for embedded training (U.S. Army Training and Doctrine Command, 2003). Embedded training, or training

that is integrated into and provided by the vehicle, can offer drivers the opportunity for training, guided practice, and

coaching at any point in their use of the vehicle. As technologies continue to develop, embedded training may

become an efficient solution to training needs.

Cross References

Developing and Maintaining Driver Mental Models, 3-10

Topic References

1. Brock, J. F., McFann, J., Inderbitzen, R. E., & Bergoffen, G. (2007). Effectiveness of commercial motor vehicle driver training curricula and delivery

methods (CTBSSP Synthesis MC-13). Washington, DC: National Academy Press.

2. Panou, M. C., Bekiaris, E. D., & Touliou, A. A. (2010). ADAS module in driving simulation for training young drivers. Proceedings of the 13th

International IEEE Annual Conference on Intelligent Transportation Systems, 1582-1587.

3. Boot, W. R., Basak, C., Erickson, K. I., Neider, M., Simons, D. J., Fabiani, M., . . . Kramer, A. F. (2010). Transfer of skill engendered by complex task

training under conditions of variable priority. Acta Psychologia, 135, 349-357.

4. Wickens, C. D., Hutchins, S., Carolan, T., & Cumming, J. (2013). Effectiveness of part-task training and increasing-difficulty training strategies: A

meta-analysis approach. Human Factors, 55, 461-470.

5. Ivancic, K., & Hesketh, B. (2000). Learning from errors in a driving simulation: Effects on driving skill and self-confidence. Ergonomics, 43(12), 1966-

1984.

6. Graving, J. S., Easterlund, P. A., & Manser, M. M. (2011). Developing a bus driver training program for a driver assistive system. Proceedings of the

Human Factors and Ergonomics Society 55th Annual Meeting, 1543-1547.

7. McKnight, J., & Adams, B. (1970). Driver education and task analysis, Volume 1: Task descriptions (Report No. DOT HS 800 367). Washington DC:

National Highway Traffic Safety Administration.

8. Rasmussen, J. (1983). Skills, rules, and knowledge: Signals, signs, and symbols, and other distinctions in human performance models. IEEE

Transactions on Systems, Man, and Cybernetics, SMC-13(3), 257-266. doi: 10.1109/TSMC.1983.6313160

9. Porter, M. M. (2013). Older driver training using video and global positioning system technology—a randomized controlled trial. Journals of

Gerontology Series A: Biomedical Sciences and Medical Sciences, 68(5), 574-580. doi:10.1093/Gerona/gls160. Epub 2012 Aug 9

10. U. S. Army Training and Doctrine Command. (2003). Objective force embedded training (OFET): User’s functional description (TRADOC Pamphlet

No. 350-37). Fort Monroe, VA: Author.

Incorporating Etiquette Into the Design of Automated Systems

Introduction

This topic provides information on presenting the automation to the driver in a manner that may support a driver’s

calibrated trust of the automated system. A strategy, often referred to as etiquette, is a set of heuristics that may

enhance and improve a driver’s interaction with the system, possibly leading to higher levels of acceptance and

performance.

Design Goal (L2, L3): Provide driver-automation interactions that are interactive and collaborative in nature.

Supporting Design Guidance

Below is a list of considerations for designing automation that is collaborative and may support the driver-automation

interaction process. These considerations were adapted from work by Lee and Seppelt (2012), itself based on Miller

and Funk (2001). The “Corresponding Topics” column refers to a topic in this document that can provide more detail

on the consideration for designing with an etiquette approach. Considerations without a corresponding topic will be

discussed further in the Discussion section of this topic.

Considerations for Designing with an Etiquette Approach Corresponding Topics

a) Make it very easy for the driver to override the automation. Communicating Transfer of Control From

System to Driver (3-8)

b) Avoid enabling interaction features just because they are

possible.

Special Considerations for Level 2 and Level 3

Automation (3-12)

c) When appropriate, explain what is being done and why. Current Automation Mode and Status (3-2)

d) When appropriate, allow driver selection of what functions

are automated.

See Discussion (next page)

e) Be aware of what the driver is doing and avoid repeating or

providing gratuitous and unnecessary messages.

See Discussion (next page)

f) Use multiple modalities to communicate. Suitable Display Properties for Automation

Mode and Status Messages (3-4);

Multimodal Messages (4-8)

g) Be casual or informal only to the extent that it furthers the

driver-automation interaction.

See Discussion (next page)

Discussion

Using an etiquette approach in automation has been shown to be beneficial for human-automation interactions and

user task performance in various domains and various forms. The use of an etiquette approach can affect a user’s trust

or reliance on the automation system (Spain & Madhavan, 2009). The considerations listed on the previous page can

be followed to design an automation system with an etiquette approach. However, every system is different and this

is not a prescriptive approach to the driver-automation interaction. In all cases, designers should consider using the

approach that best fits the scenario (Miller & Funk, 2001). Below, the considerations without a corresponding topic

are discussed.

When appropriate, allow driver selection of what functions are automated: To facilitate a collaborative relationship

between the user and the automation system, the automation system should be able to take instruction from the user

about the activities it should be engaged in Miller and Funk (2001). Specifically, this means that the user should be

able to change the function allocation between him/herself and the automation to fit the current situation without

interference from the automation system.

Be aware of what the driver is doing and do not repeat or provide gratuitous and unnecessary messages: An example

of this etiquette approach can be found in L2 and higher levels of automation in which the automation system has

communicated the need for the driver to reclaim control of the vehicle by placing his/her hands on the steering wheel.

Once there is positive indication that the driver has assumed control, the automation should not send another take

over request to the driver. If this etiquette approach is not considered, drivers may not feel comfortable trusting the

automation system because it is telling him/her information he/she already knows or is aware of, which is not

assisting the driver any further than what they are already capable of without the automation. Results of a study by

Parasuraman and Miller (2004) show that users rated the automation system as “rude” when it offered redundant,

task-specific information because they felt like the system was telling them to “hurry up.” When utilized in a military

field exercise, minimizing interruptions from non-urgent messages at critical times enabled the soldier to maintain

performance on the most critical tasks (Dorneich, Ververs, Mathan, Whitlow, & Hayes, 2012).

Be casual or informal only to the extent that it furthers the driver-automation interaction: When introducing a new

form of guided interaction, developers have sometimes tried to be informal in the interaction format, such as the use

of the anthropomorphized Microsoft Office Assistant (Horvitz, 1999). While helpful in some domains, driving is a

serious, safety-critical activity and it is important to show a clear benefit to being casual, natural, or informal before

introducing unnecessary embellishments and extra information.

Cross References

Current Automation Mode and Status, 3-2; Suitable Display Properties for Automation Mode and Status Messages,

3-4; Communicating Transfer of Control from System to Driver, 3-8; Special Considerations for Level 2 and Level 3

Automation, 3-12; Multimodal Messages, 4-8

Topic References

1. Lee, J. D., & Seppelt, B. D. (2012). Human factors and ergonomics in automation design. In G. Salvendy (Ed.), Handbook of Human Factors

and Ergonomics, Fourth Edition (pp.1615-1642). Hoboken, NJ: Wiley & Sons.

2. Miller, C. A., & Funk, H. B. (2001). Associated with etiquette: Meta-communication to make human-automation interaction more natural,

productive, and polite. Proceedings of the 8th European Conference on Cognitive Science Approaches to Process Control. Munich, Germany.

3. Spain, R. D., & Madhavan, P. (2009). The role of automation etiquette and pedigree in trust and dependence. Proceedings of the Human

Factors and Ergonomics Society 53rd Annual Meeting, 339-343.

4. Parasuraman, R., & Miller, C.A. (2004). Trust and etiquette in high-criticality automated systems. Communications of the ACM, 47(4), 51-55.

5. Dorneich, M. C., Ververs, P. M., Mathan, S., Whitlow, S., & Hayes, C.C. (2012). Considering etiquette in the design of an adaptive system.

Journal of Cognitive Engineering and Decision Making, 6 (2), 243-265.

6. Horvitz, E. (1999). Principles of mixed-initiative user interfaces. Proceedings of CHI ’99, ACM SIG-CHI Conference on Human Factors in

Computing Systems, Pittsburgh, PA.

Chapter 4. Message Characteristics

This chapter provides design guidance for developing messages that are effective for both safety

and non-safety applications. For safety messages, in particular for situations where the driver is

not actively monitoring the forward roadway or driving situation (as in L3 automation), drivers

need to respond quickly at a time when the driving situation could potentially require high

cognitive demand. Research suggests that messages presented under these conditions are most

effective when they capture drivers’ attention without being distracting, are clear and easily

understood, aid the driver in focusing attention on the roadway and/or the potential hazard, and

support the driver in making an appropriate response. In addition to safety messages, it is

expected that non-critical messages may also be presented on integrated displays. Central displays

in passenger vehicles may include integrated “infotainment” and navigation applications. The

guidelines in this chapter discuss the issues associated with DVI messages and how to present

them in a way that will enhance safety and optimize driver responses.

The following topics are included in this chapter.

• Designing Messages for Driver Comprehension

• Message Complexity

• Selection of Sensory Modality

• Multimodal Messages

Designing Messages for Driver Comprehension

Introduction

This topic provides guidance for enhancing the comprehensibility of messages. Comprehension refers to the

perceptual and cognitive processes by which drivers interpret the meaning of a message presented through a DVI.

Broader conceptual frameworks that address the comprehension of in-vehicle messages are scarce, but such a process

for design and evaluation is described by Campbell, Richman, Carney, and Lee (2004), which can give a foundation

for this topic.

Design Goal (L2, L3): Develop and present messages that support accurate and timely comprehension by the

driver.

Supporting Design Guidance

The best available research on this topic suggests that this design goal can be met if the following questions are

considered when designing messages:

Design considerations for each stage of message comprehension.

Stage of Message

Comprehension

Key Design Parameters to Consider

Extraction*

Visual messages: Character or symbol height, font, character height-to-

width/stroke-width/spacing ratios, luminance and luminance uniformity,

contrast, color, text labels (for icons and symbols).

Auditory messages: Sound level, display type, loudness, fundamental

frequencies, pitch.

Haptic messages: Type, location, amplitude/intensity/frequency.

Recognition

Temporal characteristics, level of realism and detail (for icons &

symbols), flash rate.

Interpretation

Use of color, cues to relative urgency, cues to external locations

(e.g., sound localization), use of combined cues/messages (e.g., an

auditory tone that accompanies a visual alert).

*Legibility for visual messages, complete and accurate perception of auditory and haptic messages.

Discussion

Developing effective in-vehicle messages requires a conceptual approach that applies a theoretical understanding of

driver perception and performance. As discussed in Campbell, Richman, Carney, and Lee (2004), there are three

stages associated with message comprehension and use: extraction, recognition, and interpretation. Extraction reflects

the relationships among the driver, the message, and the environment, and is essential for a complete and accurate

perception of the message. Recognition reflects the relationships among the driver, the message, and other messages

or message elements. Interpretation reflects the relationships among the driver, the message, and the referent or

underlying meaning associated with the message.

Also, as a result of familiarity, comprehension will be more likely when internationally agreed-upon icons, symbols,

words, acronyms, and abbreviations are used in the DVI (Alliance of Automobile Manufacturers, 2006)..

Design Issues

Factors affecting comprehension: Message comprehension is affected by several factors including: semantic

organization and complexity of the message, the context in which the message is presented, drivers’ familiarity with

the message, driver expectations and experience, memory limits, and workload. Familiarity has been found to be

strongly associated with comprehension. A study reported by NHTSA (2011) found that drivers responded more

slowly to a driving event (e.g., forward collision event) when they experienced an unfamiliar auditory alert. In some

cases, this difficulty remained even after reading the user manual. In short, after exposure and extended use, drivers

can learn to comprehend virtually any message. While even “bad” messages can eventually be effective, however,

they may promote errors, require training, or involve extensive trial-and-error learning.

Testing comprehension: Comprehension tests are evaluation techniques that provide a means to determine whether a

candidate message design is likely to be properly understood by typical roadway users. Several procedures can be

used to measure driver comprehension of messages, including SAE J2830 Process for Comprehension Testing of In-

vehicle Icons, which is an SAE Information Report within the SAE Standards series (SAE J2830, 2008). Given the

possible complexity of in-vehicle messages and the real possibility of multiple safety systems within a vehicle that

can present safety-critical information to the driver, it is also necessary to evaluate integrated warning systems.

Cullinane and Kirn (2012) describe a laboratory methodology that can help identify comprehension/distinguishability

issues prior to full system development in a controlled, repeatable, and safe setting.

Cross References

Developing Driver Training Material for Automated Driving System Applications, 3-14; Message Complexity, 4-4

Topic References

1. Campbell, J. L., Richman, J. B., Carney, C., & Lee, J. D. (2004). In-vehicle display icons and other information elements, Volume I:

Guidelines (Report No. FHWA-RD-03-065). McLean, VA: Federal Highway Administration. Retrieved from

www.fhwa.dot.gov/publications/research/safety/03065/index.cfm

2. Alliance of Automobile Manufacturers (AAM). (2006). Statement of principles, criteria and verification procedures on driver interactions

with advanced in-vehicle information and communication systems, including 2006 updated sections [Report of the Driver Focus-Telematics

Working Group]. Retrieved from www.autoalliance.org/index.cfm?objectid=D6819130-B985-11E1-9E4C000C296BA163

3. National Highway Traffic Safety Administration. (2011). Crash warning interface metrics, Phase 2 (Traffic Safety Facts Vehicle Safety

Research Note. Report No. DOT HS 811 471). Washington, DC: Author.

4. SAE J2830. (2008). Process for comprehension testing of in-vehicle icons. Warrendale, PA: SAE International.

5. Cullinane, B., & Kirn, C. (2012). Timely comprehension: A methodology for evaluating integration of in-vehicle warning systems. Presented

at SAE 2012 World Congress & Exhibition. doi:10.4271/2012-01-0100

Message Complexity

Introduction

This topic provides information on design characteristics that affect the complexity of messages. Specifically, it

provides a discussion of driver needs associated with message complexity and identifies characteristics of visual,

auditory, and haptic messages that affect complexity. While there is limited research specifically examining this topic

in L2 and L3 automated driving systems, designing interfaces with an understanding of well-established guidance

from past DVI research is advisable. Message complexity refers to the quantity and variety of basic information

elements contained within a message, as well as the relationships between these elements. Message complexity is an

important topic in DVI design, as messages that are too complex may not be properly perceived, comprehended, or

acted upon by the driver. The information in this topic can be used by designers to determine the level of complexity

that is appropriate for a DVI message and to implement DVI messages that are appropriately complex.

Design Goal (L2, L3): Present messages to the driver in the simplest form possible so the driver can readily

perceive, comprehend, and act upon the information.

Supporting Design Guidance

The best available research on this topic suggests that this design goal can be met when:

In General, information is presented in as simple of a manner as possible while ensuring messages

support and add value for the driver.

Visual Messages consist of simple icons and fonts with only the necessary detail included. In text

displays, the number of lines of text per-message is minimized.

Auditory Messages are simple when an immediate response is required. This could be single or

grouped frequencies presented simultaneously; such as a simple tone that consists of a square wave.

Haptic Messages are simple and perceptible. Research relevant to the topic of haptic message

complexity is limited.

Number of glances as a function of message complexity (McDougall, Tyrer, & Folkard, 2006).

Data recreated from Hoffman, Lee, McGehee, Marcias, and Gellatly (2005). Visual sampling of in-vehicle text messages: Effects of number of

lines, page presentation, and message. In Transportation Research Record: Journal of the Transportation Research Board, No. 1937, Figure 2, p. 25

and Figure 3, p. 26. Copyright, National Academy of Sciences, Washington, DC, 2005. Reproduced with permission of the Transportation

Research Board.

Automatic

Manual

Mea n Number of Glances

Line/Page

1 Line

Line Page

2 Lines

Line Page

4 Lines

Automatic

Manual

Mean Glance Duration(s)

Line/Page

1 Line

Line

Page

2 Lines

Line

Page

4 Lines

Discussion

Complexity in DVI messages generally refers to the amount of information provided in the message, but must also

include consideration of how the information will be used by the driver as well as the value of the information to the

driver.

Overall, the consequences of presenting DVI messages to the driver that are too complex can include:

disruption of attention toward the driving task (potentially an issue during transitions between automated and manual

driving), increased eyes-off-road time, increased driver workload and possible distraction, and increased response

time to critical road events.

Complexity in visual messages: Increasing the complexity of DVI messages increases cognitive demand. In Hoffman

et al. (Hoffman, Lee, McGehee, Marcias, & Gellatly 2005), a medium-fidelity simulator was used to examine how

message complexity (the number of text lines of a message) influenced visual sampling behaviors. Mean glance

duration, variability of glance duration, and the number of glances greater than two seconds all increased as the

number of lines of textual messages displayed. Visual demand was especially increased when the scrolling was

manually controlled by the driver. In another study (McDougall, Tyrer, & Folkard, 2006), the speed with which

participants searched icon arrays for a target was slower when icons were visually complex and when information

features in icons were not grouped together to form a single object. In general, icons should be simple, with only the

necessary detail included. Excessive and unnecessary amounts of detail contribute to clutter and can lead to slower

and poorer comprehension (see also Campbell, Richman, Carney, & Lee, 2004, and Easterby, 2007). These findings

are consistent with those from basic research going back at least 40 years that investigated—for example—reading

performance as a function of various characteristics of the visual stimuli (Rayner, Pollatsek, & Alexander, 2005).

Complexity in auditory messages: Simple tones are good for gaining the attention of the driver and, if properly

implemented, can be used effectively to warn of an imminent danger. Simple tones have also been shown to produce

shorter reaction times than speech messages when used in conjunction with a visual display (Kiefer, LeBlanc,

Palmer, Salinger, Deering, & Shulman, 1999). A recommendation from the CityMobil project was to limit auditory

message length to 3 or 4 information units, with messages requiring immediate reactions being as short as possible

(Martens et al., 2008). Based on the limited research in the topic, it is prudent to rely upon prior research findings and

guidance (Kiefer, LeBlanc, Palmer, Salinger, Deering, & Shulman, 1999) and limit complex auditory messages to

non-time critical situations.

Complexity in haptic messages:

Research on haptic displays in automobiles is relatively recent and has focused on

understanding what makes haptic displays perceivable to the driver and which kind of haptic warning is most

compatible with the driver response appropriate for the driving hazard (see also Lerner, Kotwal, Lyons, & Gardner-

Bonneau, 1996).

Design Issues

Reducing unnecessary message complexity is challenging in situations where multiple forms of vehicle automation

capabilities operating at multiple automation levels (e.g., L2 highway speed driving assist and L3 traffic jam assist)

are included.

Inadvertent increases in message complexity can occur in the vehicle as automation capabilities are included without

integrating the DVI components in a way that supports safe driver behaviors. In the context of discussing some

challenges to building and maintaining situational awareness in complex systems, Endsley (2012) discusses the

problem of “complexity creep,” referring to the practice of adding features and capabilities over time to systems, and

how such practices complexity can reduce the interpretability of information, reduce the predictability of the system,

and slow response time.

Cross References

Designing Messages for Driver Comprehension, 4-2; Selecting Character Height for Icons and Text, 5-10

Topic References

1. Hoffman, J. D., Lee, J. D., McGehee, D. V., Macias, M., & Gellatly, A. W. (2005). Visual sampling of in-vehicle text messages: Effects of number of

lines, page presentation, and message control. Transportation Research Record: Journal of the Transportation Research Board, 1937, 22-30.

2. McDougall, S., Tyrer, V., & Folkard, S. (2006). Searching for signs, symbols, and icons: Effects of time of day, visual complexity, and grouping.

Journal of Experimental Psychology, 12(2), 118-128. doi: 10.1037/1076-898X.12.2.118

3. Campbell, J. L., Richman, J. B., Carney, C., & Lee, J. D. (2004). In-vehicle display icons and other information elements, Volume I: Guidelines (Report

No. FHWA-RD-03-065). McLean, VA: Federal Highway Administration. Retrieved from

www.fhwa.dot.gov/publications/research/safety/03065/index.cfm

4. Easterby, R.S. (1970). The perception of symbols for machine displays. Ergonomics, 13(1), 149-158.

5. Kiefer, R., LeBlanc, D., Palmer, M., Salinger, J., Deering, R., & Shulman, M. (1999). Development and validation of functional definitions and

evaluation procedures for collision warning/avoidance systems (DOT HT 808 964). Washington, DC: National Highway Traffic Safety Administration.

6. Martens, M., Pauwelussen, J., Schieben, A., Flemisch, F., Merat, N., Jamson, A. H., & Caci, R. (2008). Human factors’ aspects in automated and semi-

automatic transport systems: State of the art (Report 3.2.1). Brussels: European Commission.

7. Lerner, N. D., Kotwal, B. M., Lyons, R. D., & Gardner-Bonneau, D. J. (1996, January). Preliminary human factors guidelines for crash avoidance

warning devices (Report No. DOT HS 808 342). Washington, DC: National Highway Traffic Safety Administration. Available at

https://archive.org/stream/preliminaryhuman00lern#page/n1/mode/2up

8. Endsley, M. R. (2012). Situation awareness. In G. Salvendy (Ed.), Handbook of Human Factors and Ergonomics, Fourth Edition (Chapter 19).

Hoboken, NJ: John Wiley & Sons.

9. Rayner; K., Pollatsek, J., & Alexander, B. (2005). Eye movements during reading. In M. J. Snowling & C. Hulme (Eds.), The Science of Reading: A

Handbook (pp. 79–97). Oxford, UK: Blackwell Publishing. doi: 10.1002/9780470757642.ch5

Selection of Sensory Modality

Introduction

This topic provides heuristics and a discussion of relevant literature to support the selection of sensory modalities

(i.e., visual, auditory, or haptic) for presenting messages in the vehicle environment. The mode of presentation can

influence driver responses and behavior. The type of modality that is appropriate for a message depends on the

driving environment (e.g., expected vehicle/cab noise and vibration, hazard scenario), the criticality of the message

(e.g., hazard vs. non-hazard situations), location of visual displays, and other factors. Further, based on current

industry trends, it is expected that many messages in L2 or L3 automated systems will be multimodal (presenting an

auditory or haptic message accompanied by a visual message). In general, much more research and analyses are

available on visual and auditory messages than on haptic. Also, haptic messages share many of the advantages and

limitations as auditory messages. This topic provides information that will help designers determine which

presentation modes are most appropriate for various messages.

Design Goal (L2, L3): Match the modality of messages with driver tasks, needs, and expectations in order to

enhance drivers’ comprehension and performance.

Supporting Design Guidance

The best available research on this topic suggests the following:

Visual messages are best for presenting more complex information that is non-safety-critical and does not call for

immediate action, and can be used to:

Provide continuous lower-priority information such as automation status, navigation-related, or

cautionary information.

Provide spatial information. In this regard, head-up displays (HUDs) and high head-down displays

(HHDDs) also have potential for presenting critical information, especially if the message has a

spatial component. (e.g., location in space relative to the driver’s vehicle).

Provide redundant or supplemental information that accompanies auditory or haptic messages.

Auditory messages are capable of quickly capturing the driver’s attention and can be used to:

Present short, simple messages requiring quick or immediate action.

Present high priority alerts and warnings (e.g., take control messages); in this instance, can be used

in conjunction with visual or haptic messages to provide redundant cues to the driver.

In an L2 system, provide an important message to drivers in situations in which they may be

distracted or looking away from a visual display (note: this may apply to haptic messages as well).

Indicate the onset of a system notification, malfunction, or limitation.

Haptic messages are capable of quickly capturing the driver’s attention and can be used if:

It is likely that the driver is in contact with the haptic display source (e.g., drivers will usually feel

a seat vibration, but will likely not be in contact with the pedal assembly or steering wheel during

L2 or L3 automated operation).

Discussion

Much has been written on the selection of visual versus auditory modes for various types of driving information and

signals. Many authors have relied on the original work of Deatherage (1972), who laid out a series of useful rules for

assisting designers in this task. The table below lists the original eight rules providing guidance for the selection of

auditory and visual mode presentations.

General rules for selecting auditory vs. visual display modes (Deatherage, 1972).

Use Auditory When…

Use Visual When…

The message is simple.

The message is complex.

The message is short.

The message is long.

The message will not be referred to later.

The message will be referred to later.

The message deals with events in time.

The message deals with locations in space.

The message calls for immediate action.

The message does not call for immediate action.

The visual system is overburdened.

The auditory system is overburdened.

The receiving location is too bright or dark.

The receiving location is too noisy.

The user must move about.

The user can stay in one place.

Most research on automotive displays focus on presenting safety-critical messages. While this is applicable for

presenting safety-critical information in automated driving systems (particularly in L2 operations), research is needed

to understand how to optimally present non-safety critical information.

The advantage of auditory and haptic displays is that they are omnidirectional signals that can command attention

regardless of where the driver is looking. In a simulator study, Stanley (2006) examined haptic, auditory, and

multimodal (haptic and auditory) displays for a lane departure warning (LDW) system. Drivers in the haptic modality

had faster reaction times than either auditory or multimodal modalities, and perceived the haptic display to be less

annoying. However, driver trust and overall preference was highest for the multimodal display. Similar findings have

been reported in other simulator studies (Scott & Gray, 2008). Likewise, a study of L2 vehicles found that drivers

both reacted and took control of the vehicle faster when receiving multimodal (visual and haptic) alerts as compared

to visual-only alerts (Blanco et al., 2015; see Chapter 2). Further research is needed to understand the implications of

using haptic displays in automotive settings, especially in non-safety critical displays.

Design Issues

Most of the relevant literature (Kiefer et al., 1999; ISO [ISO/TR 16352}, 2005; Campbell, Richman, Carney, & Lee,

2004) suggests that operator performance can be improved by combining auditory and visual messages when

presenting warnings. In addition to the above, Williges and Williges (1982) have pointed out another advantage of

visual versus auditory presentation. That is, that a visual message can be referred to until it is understood and

“encoded,” not simply referred to again later to aid with memory; an auditory signal, in contrast, is heard once

(typically), and if it is not comprehended at that time, there is not a second chance for encoding.

Cross References

Multimodal Messages, 4-8; Chapter 5: Visual Interfaces; Chapter 6: Auditory Interfaces; Chapter 7: Haptic

Interfaces

Topic References

1. Deatherage, B. H. (1972). Auditory and other sensory forms of information presentation. In H. P. Van Cott & R. G. Kinkade (Eds.), Human

engineering guide to equipment design (rev. ed.) (pp. 123-160). Washington, DC: U.S. Government Printing Office.

2. Stanley, L. M. (2006). Haptic and auditory cues for lane departure warnings. Proceedings of the Human Factors and Ergonomics Society

Annual Meeting, 50, 2405-2408. doi: 10.1177/154193120605002212

3. Scott, J. J., & Gray, R. (2008). A comparison of tactile, visual, and auditory warnings for rear-end collision prevention in simulated driving.

Human Factors, 50(2), 264-275. doi: 10.1518/001872008X250674

4. Blanco, M., Atwood, J., Vasquez, H. M., Trimble, T. E., Fitchett, V. L., Radlbeck, J., ... Morgan, J. F. (2015, August). Human factors

evaluation of level 2 and level 3 automated driving concepts (Report No. DOT HS 812 182). Washington, DC: National Highway Traffic

Safety Administration. Available at www.nhtsa.gov/sites/nhtsa.dot.gov/files/812182_humanfactorseval-l2l3-automdrivingconcepts.pdf

5. Kiefer, R., LeBlanc, D., Palmer, M., Salinger, J., Deering, R., & Shulman, M. (1999). Development and validation of functional definitions

and evaluation procedures for collision warning/avoidance systems (Report No. DOT HT 808 964). Washington, DC: National Highway

Traffic Safety Administration.

6. International Organization for Standardization. (2005). Road vehicles—Ergonomic aspects of in-vehicle presentation for transport information

and control systems—Warning systems (ISO/TR 16352). Geneva: Author.

7. Campbell, J. L., Richman, J. B., Carney, C., & Lee, J. D. (2004). In-vehicle display icons and other information elements, Volume I:

Guidelines (Report No. FHWA-RD-03-065). McLean, VA: Federal Highway Administration.

8. Williges, B. H., & Williges, R. C. (1982). Structured human/computer dialogue using speech technology. Proceedings of the Workshop on

Standardization for Speech I/O Technology (pp. 143-151). Gaithersburg, MD: National Bureau of Standards.

Multimodal Messages

Introduction

A multimodal DVI consists of more than one type of signal from the visual, haptic, and auditory modalities and is

quite suitable for delivering messages to drivers. This topic provides information on how to create multimodal driver

messages.

Design Goal (L2, L3): Present simultaneous auditory, haptic, or visual signals that enhance driver messages

and are quickly and reliably detected by the driver.

Supporting Design Guidance

The best available research on this topic suggests that this design goal can be met when:

Safety-critical messages are multimodal (either bimodal or trimodal), and consist of an effective

combination of auditory, haptic, or visual signals.

Non-safety critical messages are presented in a multimodal fashion in order to increase the likelihood

of the driver receiving the message.

Multiple, simultaneously activated, signals are used to enhance temporal redundancy, maximizing the

likelihood a driver will receive an alert (Fricke, 2007; Hecht & Reiner, 2009).

If used as an alert, visual components of a multimodal message can persist beyond the duration of

auditory or haptic signals to provide post-alert information regarding the nature of the alert (ISO,

2005).

A sequential change in modality is used for different stages of a graded system—e.g., an early

auditory tone alert to call driver attention to a less invasive signal like a visual display.

Example considerations for specific functions of different display types used for a

multimodal display that deliver imminent warning messages.

Modality

Description

Visual

Displays

Head-up Display (HUD): Used in conjunction with auditory or haptic alerts to encourage drivers to

attend to the forward roadway in safety critical situations (Lind, 2007); safety concerns may arise if HUD

images block the driver’s view of forward hazards.

High Head-down Display (HHDD): Used to improve noticeability of the visual aspects of a message for

drivers who are unable to hear or feel the signals from the other modalities that are used (ISO, 2005).

Low Head-down Display (LHDD): Used to present visual messages; these are paired with auditory or

haptic signals (ISO, 2005). For safety critical messages, the presentation period may begin after the

warning criterion is no longer exceeded, following an auditory or haptic signal (ISO, 2005).

Instrument Panel (IP) Display: Used to present visual messages. However, the use of this space as part

of a multimodal display for safety critical information is not recommended (Perez, Kiefer, Haskins, &

Hankey, 2009; SAE J2400, 2003).

Auditory

Displays

Speech Messages: Speech can also be used to supplement or more clearly indicate information provided

in a visual or haptic display, such as providing messages or instruction to the driver.

Simple Tones (Conventional Auditory): Simple tones are commonly paired with haptic and visual.

Spatial messages using simple tones are enhanced greatly when coupled with spatial haptic messages

(e.g., Fitch, Kiefer, Hankey, & Kleiner, 2007).

Haptic

Displays

Vibrotactile Seat: Commonly paired with auditory or visual displays, e.g., the addition of audio or visual

displays may facilitate driver comprehension of more complex vibrotactile seat displays.

Other Haptic/Tactile Displays: May be implemented as part of a multimodal warning system. However,

limited research is available regarding the use of seat belt pre-tensioning or brake pulses as a component

of a multimodal display.

Discussion

Many vehicle displays are bimodal and can present both visual and auditory signals. A bimodal presentation scheme

can be employed for in-vehicle tell-tale messages using a tone, to alert the driver to a system issue, coupled with a

visual component (e.g., the tell-tale visual icon) to convey the nature of the system problem (ISO, 2005). The use of

multimodal displays may also help ensure messages are delivered to drivers with unisensory deficits (e.g., older

populations; see Laurienti, Burdette, Maldjian, & Wallace, 2006) or drivers who are not engaged with the driving

task (e.g., distracted or operating in highly automated conditions).

Auditory and haptic signals are beneficial when used to deliver a message for a multimodal display because they are

detectable even when drivers are glancing away from the roadway or information displays. A study comparing

visual-only to combination visual and haptic messages in L2 vehicles demonstrated that drivers reacted and took

control of the vehicle faster when receiving multimodal take control messages (Blanco et al., 2015, see Chapter 2).

Visual messages can serve as “back-up” communication channels if there is high ambient noise/vibration or if a

driver is hearing impaired, as well as a method for portraying the nature of the alert (ISO, 2005).

The reference to Fitch, Kiefer, Hankey, and Kleiner (2007) is included because their results were obtained using

laboratory methods that have broader application. They investigated a multimodal presentation format for providing

directional alerts for collision avoidance systems and had their participants simply indicate the direction participants’

felt the interface was indicating.

Design Issues

Correspondence and Redundancy: Forming a full correspondence between stimuli across multiple modalities that are

not normally associated with one another is unlikely to occur without significant training (Spence, Ngo, Lee, & Tan,

2010) or experience. Therefore, messages need to be intuitive and easily learned so that drivers can straightforwardly

begin to integrate multimodal signals over time. For higher priority messages, early timing is recommended to give

drivers adequate time for processing and responding to multimodal displays that deliver complex messages (e.g.,

spatial cues). Redundant temporal cues (e.g., simultaneous signals at the same time) may be adequate to pique the

attention of a driver.

Cross References

Developing Driver Training Material for Automated Driving System Applications, 3-14

Topic References

1. Fricke, N. (2007). Effects of adaptive information presentation. Proceedings of the 4th International Driving Symposium on Human Factors in

Driver Assessment, Training and Vehicle Design, 292-298.

2. Hecht, D., & Reiner, M. (2009). Sensory dominance in combinations of audio, visual and haptic stimuli. Experimental Brain Research, 193,

307-314.

3. International Organization for Standardization. (2005). Road vehicles—Ergonomic aspects of in-vehicle presentation for transport information

and control systems—Warning systems (ISO/TR 16352). Geneva: Author.

4. Lind, H. (2007). An efficient visual forward collision warning display for vehicles (SAE Technical Paper 2007-01-1105). Warrendale, PA:

SAE International. doi:10.4271/2007-01-1105.

5. Perez, M. A., Kiefer, R. J., Haskins, A., & Hankey, J. M. (2009). Evaluation of forward collision warning system visual alert candidates and

SAE J2400. SAE International Journal of Passenger Cars- Mechanical Systems, 2(1), 750-764. doi:10.4271/2009-01-0547

6. SAE J2400 (2003). Human factors in forward collision warning systems: Operating characteristics and user interface requirements.

Warrendale, PA: SAE International.

7. Fitch, G. M., Kiefer, R. J., Hankey, J. M., & Kleiner, B. M. (2007). Toward developing an approach for alerting drivers to the direction of a

crash threat. Human Factors, 49(4), 710-720.

8. Laurienti, P. J., Burdette, J. H., Maldjian, J. A., & Wallace, M. T. (2006). Enhanced multisensory integration in older adults. Neurobiology of

Aging, 27, 1155-1163.

9. Blanco, M., Atwood, J., Vasquez, H. M., Trimble, T. E., Fitchett, V. L., Radlbeck, J., ... Morgan, J. F. (2015, August). Human factors

evaluation of level 2 and level 3 automated driving concepts (Report No. DOT HS 812 182). Washington, DC: National Highway Traffic

Safety Administration. Available at www.nhtsa.gov/sites/nhtsa.dot.gov/files/812182_humanfactorseval-l2l3-automdrivingconcepts.pdf

10. Spence, C., Ngo, M. K., Lee, J.-H., & Tan, H. (2010). Solving the correspondence problem in haptic/multisensory interface design. In M. H.

Zadeh (Ed.). Advances in Haptics (pp. 47-74). Available from: www.intechopen.com/books/advances-in-haptics/solving-thecorrespondence-

problem-in-haptic-multisensory-interface-design

Chapter 5. Visual Interfaces

This chapter contains topics on visual displays. The visual modality is of primary importance in

the driving task, and is amenable to the use of various sensory dimensions such as color,

luminance and contrast, as well as stimulus dimensions such as location, size and shape and

periodicity (e.g., flashing). Additionally, vision is the channel for presenting written information,

and so is appropriate for messages from the vehicle automation involving semantic content that

benefits from persistence, as distinct from auditory linguistic warnings, which tend to be obtrusive

if they persist.

While vision provides a rich field for information coding and potentially complex messages that

can help interpret messages provided by the DVI of automated driving systems, some challenges

must be addressed in order to ensure their effectiveness. Visual warnings must be seen to be

effective, and placing them in optimal locations in the cab can facilitate rapid detection of visual

signals and promote faster responses to them. In addition, characteristics such as display type,

color, size, spacing, and temporal characteristics (e.g., flashing or apparent motion) can be chosen

to maximize the conspicuity, legibility, and comprehensibility of warning messages.

Glare from strong light sources presents another challenge to visibility, conspicuity, and legibility

of warning messages. Effective warnings depend on the display having sufficient contrast that

drivers can easily detect and read the images presented thereon. Glare on the display reduces the

contrast of images presented on the display, while glare emanating from the display reduces

contrast sensitivity in the eye. Both sources of glare can potentially reduce the effectiveness of

warning displays by limiting the visibility of messages. This chapter discusses methods for

mitigating glare, both on the display and from the display, in order to prevent loss of contrast.

Head-up displays have the potential to provide drivers with critical, forward-oriented information

while minimizing glance times away from the forward roadway scene, potentially reducing eye

movement and accommodation time. Images presented on the HUD, however, have the potential

to be distracting and can partially occlude important visual cues outside the cab. Consequently,

warning displays presented on HUDs should be designed with care. Nevertheless, the HUD can

be an effective display for presenting time-critical messages.

Design topics addressed in this chapter:

• Locating a Visual Display

• Display Glare

• Head-Up Displays

• Using Color

• Selecting Character Height for Icons and Text

• Temporal Characteristics of Visual Displays

Locating a Visual Display

Introduction

The location of a visual display is a key factor affecting the ease with which drivers can obtain information. The

placement of the visual component of a DVI will facilitate access to the information. The main focus of this topic is

locations for visual displays, some attention is devoted to techniques to minimize glare for increased readability.

Design Goal (L2, L3): Place the visual interface in a location that facilitates rapid extraction of information

while minimizing eyes-off-road glances and negative impacts on driving performance.

Supporting Design Guidance

The best available research on this topic suggests that this design goal can be met when:

Critical displays for continuous vehicle control or critical warnings related to vehicle forward path are

within ±15 degrees of the central line of sight but as close to the central line of sight as practicable.

ISO (1984) recommends that messages that require immediate detection be located within 5 degrees

of the forward view when possible.

Displays are designed and located to minimize glare from external sources or other displays in the

vehicle (e.g., in the instrument panel or under a protective cover.

Examples of potential visual display locations (indicated by red squares and callouts).

Display Locations in Image:

A. Head-Up Display

B. High Head-Down Display

C. Head-Down Display/Instrument Panel

D. Center Console

Discussion

In general, placing a visual display near the forward line of sight to the primary driving task may increase the

likelihood that it will be seen, and will reduce the time needed to glance at that information, when drivers are looking

out of the windscreen. The example display location of 15 degrees within the line of is taken from the ComSIS

preliminary guidelines (Lerner, Kotwal, Lyons, & Gardner-Bonneau, 1996), but the relevance to L2 and L3

automated driving has yet to be validated. ISO (2005) recommends that visual displays intended to send critical

messages to the driver be located as near to the driver’s line of sight as possible, particularly if color is used. ISO

(1984) recommends that critical visual messages that require immediate detection should be located within 5 degrees

of the driver’s line of sight.

Lind (2007) compared a collision warning HUD displayed in the central driver view in response to critical road

hazards, with displays on the upper dashboard (HHDD), instrument cluster, and a steering wheel array of light

emitting diodes (LEDs). This study involved a strictly visual HUD, i.e., there was no associated auditory cue. The

salient attentional signal was flashing of the LED matrix (presenting at 4 Hz for 1.2 s). This type of HUD is different

than more conventional implementations of HUDs, which may use alphanumerics, graphics or icons.

Design Issues

Although there is limited research on locations for L2 and L3 automated driving system messages exists, it can

generally be assumed that drivers will expect key visual information to be presented in the general vicinity of the

instrument panel. However, existing guidance is based upon the assumption that the driver is generally looking

towards the forward roadway. L3 automated driving systems operate with the assumption that drivers will not be

looking towards the road [5], and extended eyes off road time is possible in L2 operations. Further research is needed

as to how drivers sample information from displays during both L2 and L3 automated driving.

Cross References

Multimodal Messages, 4-8

Topic References

1. International Organization for Standardization (ISO) (1984). Development and principles for application of public information symbols.

(ISO/TR 7239). Geneva: Author.

Lerner, N. D., Kotwal, B. M., Lyons, R. D., & Gardner-Bonneau, D. J. (1996, January). Preliminary human factors guidelines for crash avoidance

warning devices (Report No. DOT HS 808 342). Washington, DC: National Highway Traffic Safety Administration. Available at

https://archive.org/stream/preliminaryhuman00lern#page/n1/mode/2up

3. International Organization for Standardization. (2005). Road vehicles—Ergonomic aspects of in-vehicle presentation for transport information

and control systems—Warning systems (ISO/TR 16352). Geneva: Author.

4. Lind, H. (2007). An efficient visual forward collision warning display for vehicles (SAE Technical Paper 2007-01-1105). Warrendale, PA:

SAE International. doi:10.4271/2007-01-1105

5. Llaneras, R. E., Salinger, J., & Green, C. A. (2013). Human Factors issues associated with limited ability autonomous driving systems:

Drivers' allocation of visual attention to the forward roadway. Proceedings of the International Driving Symposium on Human Factors in

Driver Assessment, Training, and Vehicle Design, 7, 92-98.

Display Glare

Introduction

Glare on visual displays can originate from a variety of sources in the driving environment and can make visual

displays difficult to read. In addition, light emanating from displays can be glaring at night causing discomfort, or in

some conditions, reduced visibility of the external driving environment. This topic discusses ways to mitigate both the

reduced legibility and conspicuity of display information due to glare on the display and the reduced visibility of the

environment and increased physical discomfort caused by glare from in-vehicle displays.

Design Goal: Minimize glare, both on and from, visual displays.

Supporting Design Guidance

The best available research on this topic suggests that this design goal can be met through the following strategies:

Mitigating glare on the display in daytime driving:

Provide sufficient display luminance and use high contrast display technologies to ensure adequate

contrast.

Place safety-critical displays in a location that minimizes exposure of the display sunlight.

Use designs or locations that provide shading, such as a cowling or an inset bezel.

Use anti-glare coatings or films to filter incoming light and reduce glaring reflections from the

display.

In some configurations, smaller display sizes can be easier to shade; however, care must be taken to

ensure that other important design considerations, such as symbol size and conspicuity, are not

compromised.

Mitigating glare that emanates from the display while driving in darkness:

Provide a control that allows drivers to adjust the display intensity but do not allow drivers to turn the

display off completely.

Use light sensors to automatically reduce display luminance in darkness.

Display content using a dark background to minimize the luminance emanating from the display.

Locate and orient the display to minimize reflections on windows.

Consider locating non-safety-critical displays in highly eccentric locations relative to the forward

gaze (e.g., center stack) to increase the glare angle. Do not use this approach for critical safety

messages.

Examples of mitigations for glare incident on and emanating from the display.

Glare on the Display

in Daylight

A. Display embedded

in the instrument

panel to protect

from sunlight

B. Display mounted in

recess above the

center stack

A & B. Display

luminance is

sufficient to ensure

adequate contrast

Glare from the Display

in Darkness

C. Display intensity is

adjustable to allow

drivers to control

amount of luminance

The CIE veiling luminance model to

the left shows that veiling

luminance (



) increases as (1)

glare intensity (



) increases and

(2) glare angle () decreases.

Increased veiling luminance results

in reduced visibility

Adapted from (Vos, et al., 2002). Used with permission.





= 



 





+ 





  1 + 



62.5







Where











=  





=     

 =  

 =  

Discussion

Glare on the display: Intense light, such as sunlight, that falls on a visual display superimposes a uniform luminance

onto the display, essentially “filling in” the darker areas of the displayed image, thereby reducing image contrast (i.e.,

the luminance ratio of the light to dark areas is reduced). The image on the display becomes increasingly difficult to

read as the contrast decreases, until eventually the image can no longer be detected ((Vos, et al., 2002; Human Factors

and Ergonomics Society, 2007). This reduction in contrast could be a particular problem for applications that rely on

visual displays to present time-critical safety messages because lower contrast can increase drivers’ reaction times or

they may not see the display altogether.

Glare from the display: Glare from a visual display occurs when the intensity of the display within the visual field is

substantially greater than the visual adaptation level, causing physical discomfort or pain (discomfort glare) and/or

reduced visibility (disability glare). A portion of the light entering the eye is scattered in the transparent media of the

eye (i.e., cornea, lens, and vitreous fluids) and by the tissues in the ocular fundus (Adrian & Bhanji, 1991). Some light

also diffuses through the sclera and iris tissues. The scattered light superimposes a uniform veiling luminance onto the

retinal image, reducing its overall contrast. If the contrast of the image falls below the contrast threshold for visibility

under these conditions, it will be rendered invisible (Vos et al., 2002). Veiling luminance is influenced primarily by the

intensity of light, the surface area of the lighted areas of the display, and the angle at which the glaring luminance

enters the eye.

Design Issues

Glare on the display: Preventing glare from sunlight falling on an in-vehicle display can be extremely challenging to

designers. Kiefer et al. (1999) recommend that one way to mitigate glare on a display is to provide sufficient

luminance from the display in daytime driving to ensure adequate contrast. High-contrast display technologies can also

reduce the effects of glare on the display. (Wreggit, Powell, Kirn, & Hayes, 2000) found that an electroluminescent

display provided sufficient contrast for legibility, and drivers reported no washout or glare from sunlight. Because

vehicles are not stationary, it may be difficult (or perhaps impossible) to locate a display in a location that will never

receive direct sunlight. Nevertheless, placing the display in the instrument panel, in a custom recess, or within a

shading bezel, etc., can help reduce exposure to glaring light. In some configurations, smaller displays may be easier to

protect from direct light because they have less surface area to shade; however, it is important to ensure that other

aspects of DVI design (e.g., text and icon legibility, conspicuity) are not compromised when using a smaller display.

Glare from the display: Several mathematical models have been developed that estimate the amount of veiling

luminance developed by a glare source (e.g., Vos et al., 2002; Adrian & Bhanji, 1991; Farber & Matle, 1989). These

models show that veiling luminance is directly proportional to intensity and inversely proportional to the angle at

which glaring luminance enters the eye relative to the forward gaze. Thus, there are two primary solutions for reducing

the effects of glare emanating from in-vehicle displays: (1) reduce the amount of light emanating from the display

and/or (2) increase the eccentricity of the display location. A preferred approach for reducing the effects of glare from

displays is to provide a control that is used to adjust the display intensity This can be a manually operated control that

drivers manipulate or an automated control that adjusts display luminance based on sensors that detect ambient light

levels. Regardless of how the control is implemented, however, the amount of control provided should be limited,

however, to prevent drivers from turning off the display completely (Kiefer et al., 1999). Another way to reduce

display luminance is to present content on a dark background in order to minimize the overall surface area of high-

intensity portions of the image. Finally, locating the display further into peripheral vision can reduce the effects of

glare, but with important tradeoffs with regard to reductions in warning conspicuity and detection.

Cross References

Locating a Visual Display, 5-2

Topic References

1. Vos, J. J., Cole, B. L., Bodmann, H.W., Colombo, E., Takeuchi, T., & van den Berg, T.J.T.P. (2002). CIE equations for disability glare (CIE

146:2002). Vienna: International Commission on Illumination [CIE].

2. Human Factors and Ergonomics Society (2007). Human factors engineering of computer workstations. Santa Monica, CA: Author.

3. Adrian, W., & Bhanji, A. (1991). Fundamentals of disability glare: A formula to describe stray light in the eye as a function of glare angle and

age. Proceedings of the First International Symposium on Glare (pp. 185-193). New York: Lighting Research Institute.

4. Kiefer, R., LeBlanc, D., Palmer, M., Salinger, J., Deering, R., & Shulman, M. (1999, August). Development and validation of functional

definitions and evaluation procedures for collision warning/avoidance systems (Report No. DOT HT 808 964). Washington, DC: National

Highway Traffic Safety Administration. Available at

www.nhtsa.gov/DOT/NHTSA/NRD/Multimedia/PDFs/Crash%20Avoidance/1999/HS808964_Report-1999-08.pdf

5. Wreggit, S. S., Powell, J., Kirn, C., & Hayes, E. (2000). Human factors evaluation of the truck productivity computer’s electroluminescent

display (Paper No. 2000-01-3409). Warrendale, PA: SAE International.

6. Farber, E., & Matle, C. (1989). PCDETECT: A revised version of the DETECT seeing distance model. Transportation Research Record:

Journal of the Transportation Research Board, 1213, 11-20.

Head-Up Displays

Introduction

Head-up displays (HUDs) have the potential to provide drivers with critical information while minimizing glance

times away from the forward roadway scene. This can increase the speed of information access by the driver by

reducing eye movement and visual accommodation time. HUDs have been used and studied extensively in

aviation (Prinzel & Risser, 2004) and, as costs have reduced, are being used more frequently in automobiles. HUDs

also have the potential to expand display space within the vehicle. Application of the technology may proceed

conservatively, however, while impacts of factors such as potential distractions and driver individual differences are

further evaluated. Designers can consider the necessity of the information with regard to the current driving situation

when deciding what information to provide via HUD. However, there exists a sufficient body of information on

HUDs - both in aviation and, more recently, in automotive applications - to establish design guidance.

Design Goal (L2, L3): Use HUDs to present simple indications of critical safety situations in the driver’s

forward view.

Supporting Design Guidance

The best available research on this topic suggests that this design goal can be met when:

Information presented via HUD is interpretable within the HUD and does not require visual reference

to other head down displays.

Information relevant to the driving situation is prioritized over presentation of non-driving related

information.

The use of continuously presented stable-value information on a HUD is minimized.

The use of symbols, text, or indicators that continuously change in value is limited.

The HUD is located 5 degrees to the right and 5 degrees below the center line of driver view.

Two potential uses and implementations of HUD displays.

HUD providing a warning.

Only present during alert

Used in conjunction with

other modalities

Removed after scenario

resolved

HUD providing navigation information.

Only provides

driving/navigation relevant

information

Does not provide stable-

value information (e.g.,

“Navigation Active”)

Minimal number of

dynamic elements

Driver may disable if

desired

Discussion

While HUD use in L2 and L3 automation applications remains an unexplored topic from the standpoint of publically

available literature, guidance from automotive HUD use in safety applications are useful. These principles are

primarily applicable during L2 operations or when the driver’s attention must be reoriented due to a pending transfer

from automated to manual control.

HUDs have been studied in a number of different configurations and conditions, including comparisons with

HHDDs, traditional instrument clusters, and under various traffic load, secondary task and hazard detection

conditions. The clearest result to emerge from this work is the advantage of the HUD in reducing braking time and

increasing warning detections for critical road events (Lind, 2007; Kieffer, 1996).

The balance of evidence suggests that under test conditions, HUDs or HHDDs (located slightly lower than HUDs, yet

above the instrument cluster) tend to improve driver performance as measured by vehicle headway distance, response

time to critical events, and other measures of driving behavior (Lind, 2007; (Charissis & Papanastasiou, 2010;

Horrey, Wickens, & Alexander, 2003; Perez, Kiefer, Haskins, & Hankey, 2009).

Location of a HUD is the variable with the most consistent findings across a range of experiments, which suggest that

the HUD should be located approximately 5 degrees to the right and 5 degrees below the driver’s central visual focus

(Yoo, Tsimhoni, Wantanabe, Green, & Shah, 1999). There is some evidence that display HUDs are preferred by

drivers in simulator studies, and they yield better driving performance and information detection than a head-down

display (HDD) (Charissis & Papanastasiou, 2010).

Design Issues

As a platform for safety-critical messages, HUDs are most appropriate for providing drivers with information

relevant to the forward roadway or attracting the driver’s attention during an automated to manual transition. This

approach reserves the HUD location for critical information that is unlikely to be extraneous or distracting. The

designer should carefully consider what information is presented in the HUD and only provide information that is

relevant to the driver, minimizing presentation of non-driving related information. Presentation of stable-value

information, especially when not directly related to the driving task, should be minimized. Likewise, dynamic

elements that are not critical to the driving task or system status should not be incorporated.

Other approaches, such as providing vehicle or automation system status, are more complex and involve a much

larger range of display elements and properties. While there seems to be no documented detrimental effect of these

types of displays in laboratory and limited on-road testing for passenger vehicles, they may have the effect of

reducing the impact or salience of critical forward warnings, or worse, lead to an excessively cluttered field of view.

Cross References

Display Glare, 5-4

Topic References

1. Prinzel, L.J., III, & Risser, M. (2004). Head-up displays and attention capture (Report No. NASA/TM-2004-213000). Retrieved from the

NASA Technical Reports Serverwebsite at

http://naca.larc.nasa.gov/search.jsp?R=20040065771&qs=N%3D4294966788%2B4294724598%2B4294323833

2. Lind, H. (2007). An efficient visual forward collision warning display for vehicles (SAE Technical Paper 2007-01-1105). Warrendale, PA:

SAE International.doi:10.4271/2007-01-1105

3. Kiefer, R. J. (1996). A review of driver performance with head-up displays. Third World Congress on Intelligent Transport Systems

[CD-ROM]. Washington, DC: ITS America.

4. Charissis, V., & Papanastasiou, S. (2010). Human-machine collaboration through vehicle head up display interface. Cognition, Technology

and Work, 12(1), 41-50.

5. Horrey, W. J., Wickens, D. D., & Alexander, A. L. (2003). The effects of head-up display clutter and in-vehicle display separation on

concurrent driving performance. Proceedings of the Human Factors and Ergonomics Society 47th Annual Meeting, p. 1880-1884.

6. Perez, M. A., Kiefer, R. J., Haskins, A., & Hankey, J. M. (2009). Evaluation of forward collision warning system visual alert candidates and

SAE J2400. SAE International Journal of Passenger Cars- Mechanical Systems, 2(1), 750-764. Warrendale, PA: SAE International.

doi:10.4271/2009-01-0547

7. Yoo, H., Tsimhoni, O., Wantanabe, H., Green, P., & Shah, R. (1999). Display of HUD warnings to drivers: Determining an optimal location

(Report No. UMTRI-99-9). Ann Arbor: University of Michigan Transportation Research Institute.

Using Color

Introduction

Color is a characteristic of visual displays that can be useful for conveying meaning or urgency of alerting signals.

Color has certain advantages over text and symbols in terms of immediacy of recognition, and can serve to reinforce

meaning conveyed by other methods. Color is a complex variable, however, and issues of luminance, hue, contrast

and potential conflicts with other messages must be considered during design.

Design Goal: Use color to augment visual information presented in the displays to attract attention and/or to

convey the urgency of conditions or situations.

Supporting Design Guidance

The best available research on this topic suggests that this design goal can be met when:

The color is associated with the level of warning:

− Red is normally associated with danger or critical situations

− Yellow is normally associated with caution

− Green is normally associated with normal operation; however, other considerations about warning

conspicuity may necessitate using a different color (see Design Issues on the next page).

The colors that are used are compatible with symbols based on prior association, such as red for

octagonal stop signs, and yellow for triangular or diamond warnings.

The quantity of colors used to code information is minimize; do not to exceed 4 color codes.

Color is used to create a “pop-out” effect in forward collision warnings to show the area of concern

more distinctly from the background scene.

The following color contrast combinations are avoided: green/red, green/blue, yellow/red,

yellow/blue, violet/red.

Example illustrating a potential use of color as part of a system status display.

This image shows a hypothetical lane centering status display. The left

image indicates that the car is not centered in the lane. The non-centered

status is provided through position of the car on the lane display, as well

as through the red color of the car. The right image indicates that the car

is centered in the lane. The centered status is portrayed through showing

the car in the center of the lane display and through the green color of the

car.

Discussion

The stereotyped interpretation of certain colors can be used in combination with other information to convey or

provide messages to the driver, and, importantly, to promote appropriate and timely responses when compatible with

stereotypical stimulus-response pairings. The traditional association of the color red with “danger or critical

situation,” yellow with “caution,” and green with “normal” can be used to compliment auditory or haptic signals, and

to convey urgency (ISO, 2005). While there have been some recent findings that the association of red with “danger”

is stronger than the association of yellow with “caution” (Leonard, 1999), the context of driving would tend to

reinforce the stereotype interpretation based on the frequency with which such colors are encountered in the roadway

environment.

During development of the air bag warning label, NHTSA focus groups (49 C.F.R. pt. 571.208, 2011), did not

associate orange with the word “danger.” To reduce potential ambiguity, and to maximize perceived color

distinctiveness, yellow should be used to indicate a discrete system warning or cautionary state (however, see Design

Issues below). The color green should not be used to provide a warning to the driver because it is associated with safe

or normal operating conditions (ISO, 2005).

Color and shape combinations are used for specific types of warnings and traffic regulation on road signs, and use of

the same combinations for in-vehicle messages will help to maintain consistency between the road infrastructure and

in-vehicle information environments. For example, use of a diamond shaped symbol with yellow background and

black text is the accepted standard for warning signs (FHWA, 2012), and application of these conventions for specific

automated driving system warning messages is appropriate. Color-shape combinations can increase reaction times

when signs utilize shape as well as text because additional decision elements are used; thus if a shape is associated

with a highly stereotyped response, additional text may lead to longer reaction times (Tijus, Barcenilla, Cambon de

Lavalette, & Maunier, 2007).. Designers should avoid incompatible or unconventional shape and color combinations,

such as octagonal shape (conventional meaning = stop) presented with a yellow background (conventional meaning =

caution).

Color contrast can affect the perception of both the background and message content through complex interactions of

luminance and visual system effects. Avoidance of the specific combinations described in the Design Guidance above

will preclude this problem (ISO, 2005). For messages requiring the presentation of text, green text has the advantage

of being at the frequency of maximum spectral sensitivity of the eye (NASA, n.d.). Similarly, green-yellow text (534

nm) best accommodates both light- or dark-adapted eyes.

Design Issues

Color should not be used as the primary or exclusive means by which information is conveyed, but instead a

supplementary element or alternative cue to meaning. When approached in this way, designers can think of

appropriate location and symbol-shape means to convey the principal message content, and color can be used as a

means to more quickly draw attention and reinforce meaning through traditional associations. Furthermore, relying

on characteristics other than color will convey to drivers with color-blindness important information that might

otherwise be missed if color were used exclusively to communicate the information.

Keeping the number of color codes within human cognitive limits is important in the driving environment. Four

colors is the recommended maximum, as this corresponds generally to the warning levels of danger, warning, caution

and normal operation (ISO, 2005).

Cross References

Designing Messages for Driver Comprehension, 4-2; Locating a Visual Display, 5-2

Topic References

1. International Organization for Standardization. (2005). Road vehicles—Ergonomic aspects of in-vehicle presentation for transport information

and control systems—Warning systems (ISO/TR 16352). Geneva: Author.

2. Leonard, S. D. (1999). Does color of warning affect risk perception? International Journal of Industrial Ergonomics, 23, 499-504.

3. 49 C.F.R. pt. 571.208. (2011, October 1). Federal Motor Vehicle Safety Standard (FMVSS) No. 208, Occupant crash protection

4. Federal Highway Administration (FHWA). (2012). Manual on uniform traffic control devices, 2009 edition with revisions 1 and 2.

Washington, DC: Author. Available at https://mutcd.fhwa.dot.gov/pdfs/2009r1r2/mutcd2009r1r2edition.pdf

5. Tijus, C., Barcenilla, J., Cambon de Lavalette, B., & Maunier, J.-G. (2007). The design, understanding and usage of pictograms. In

G. Rijlaarsdam (Series Ed.) Vol. 21. Studies in Writing: Written Documents in the Workplace (pp. 17-31). Netherlands: Kluwer.

6. NASA Ames Research Center Color Usage Research Lab. (n.d.) Using color in information display graphics. Retrieved from the NASA Ames

Research Center Color Usage Research Lab website at http://colorusage.arc.nasa.gov/index.php

Selecting Character Height for Icons and Text

Introduction

Optimum text and graphic symbols presented in the driver-vehicle interface are legible under a large range of

viewing distances, viewing angles, and environmental conditions. Legibility goes beyond visibility or detection; it

implies being able to discern shape or character identity based on appearance. An important factor that influences

legibility of an icon is the size of the symbols that comprise the icon—both graphical symbols and text. This topic

addresses symbol size, which refers to the visual angles subtended by the icon, its graphical elements, text within the

icon, and free-standing text.

Design Goal (L2, L3): Select sizes for text and icons that support rapid message legibility.

Supporting Design Guidance

The best available research on this topic suggests that this design goal can be met when:

Icon Size:

Optimal visual angle of primary graphical elements

: 86 arcminutes

Minimum visual angle of primary graphical elements

: 41 arcminutes for time-critical applications

34 arcminutes for non-time-critical applications

Text Size (both within the icon and free-standing text—see the discussion):

Optimal height: 20 arcminutes

Minimum height: 16 arcminutes for time-critical applications

12 arcminutes for non-time-critical applications

Primary graphical elements provide the primary information needed to encode or detect the icon. Secondary

graphical elements provide additional context or clarifying information. Optimum visual angle refers to the angle

at which the primary graphical elements are both conspicuous and legible. Minimum visual angle refers to the

smallest angle at which the primary graphical elements are legible but not necessarily conspicuous [2].

The table below provides equations for calculating the sizes of the icon, its graphical elements, text within the icon,

and free-standing text. Note that the equations assume the visual angle is measured in arcminutes, symbol height is in

millimeters, viewing distance is in meters, and the trigonometric functions (tangent and arctangent) accept and return

values in degrees rather than in radians. Appropriate conversion factors must be applied for different units.

Equations for calculating symbol height, visual angle, and viewing distance.

If Known…

Use These Equations for Calculating These Unknowns

Visual Angle (V)

in arcminutes

Symbol Height (H)

in millimeters

Viewing Distance (D)

in meters

Viewing Distance (D) and

Symbol Height (H)

= 60  Arctan 



1000  



— —

Viewing Distance (D) and

Visual Angle (V)

—

= 1000  

Tan







—

Visual Angle (V) and

Symbol Height (H)

— —

=



1000 ∙ Tan







Definitions of Variables Used in the Equations

Figures adapted from Campbell et al. (2007)

Examples of Symbol Elements in an Icon

Eyepoint

Visual

Angle

Symbol

Height

(Distance from Viewer to Display)

H = Symbol height in millimeters

D = Viewing distance in meters (0.5–1.1 m)

V = Visual angle subtended in arcminutes

Border

Background

Secondary Graphical

Element

Text label

WARNING

Symbol

WARNING

Primary Graphical

Element

Discussion

Icon size: The design guidance above for icon size are consistent with the recommendations made by

ISO/TR7239 (ISO, 1984), which were based on a variety of research related to detection and resolution thresholds.

The optimum visual angle suggested (86 arcminutes) is aimed at ensuring conspicuity, while the minimum visual

angle (41 arcminutes) simply ensures legibility. It is important to note that the recommendations made by ISO

(1984)—and therefore the supporting design guidance on the previous page—are based on the assumption that the

icon will not be placed outside a 15-degree angular displacement from the central line of the normal direction of

user’s vision. ISO (1988) specifies that the minimum size of graphical symbols for use on equipment should be

1/100th their viewing distance, which corresponds to 34 arcminutes of visual angle. The larger sizes in ISO (1984)

are recommended for time-critical applications in order to ensure both conspicuity and legibility in the driving

environment.

Text size: ISO 15008 (ISO, 2009) recommends that character heights for in-vehicle display text should subtend at

least 20 arcminutes of visual angle, but 16 arcminutes is acceptable. Furthermore, the minimum visual angle for text

should be no less than 12 arcminutes, but text of this size should be reserved for situations with only modest

requirements for reading speed and accuracy. More recent research by O’Day and Tijerina (2011) verifies these

values. They found that the highest accurate reading rate occurred with the largest text height they tested (20

arcminutes) and the lowest accurate reading rate for text that subtended 12 arcminutes. In addition, the greatest

variability in accurate reading rate was associated with the smallest text. Taken as a whole, ANSI (HFES, 2007),

Mourant and Langolf (2007), Howell and Kraft (1959), and Giddings (1972) agree with the ISO (2009) standard,

recommending a minimum character height of 16 arcminutes and optimal character heights for high legibility in the

range of 20 to 30 arcminutes. The recommendation given in this design principle reflects the specifications for text

height found in ISO (2009) because it is an international standard that applies directly to the presentation of textual

information in vehicle-based applications.

It should be noted that the literature did not provide any information that suggested the size of text within an icon

should be different than the size of free-standing text in terms of legibility. Therefore, the design principle above does

not differentiate between these implementations.

Design Issues

The size of in-vehicle displays is often limited by the available real-estate in the cab, which in turn limits the size of

the symbols presented on the displays. This limitation can result in a tradeoff between symbol size and legibility.

When designing in-vehicle displays, it is important to consider legibility when determining the sizes of symbols,

especially in safety-critical applications where the time available to read and interpret the symbols is limited. O’Day

and Tijerina (2011) found that a wide variety of character heights can be legible if the character width and stroke

width are carefully chosen. Nevertheless, size is only one of the characteristics of graphical and textual symbols that

affect legibility. The legibility of icons and text is determined by factors such as the size, stroke width, contrast, and

luminance (Howell and Kraft, 1959; Carney, Campbell, & Mitchell, 1998).

Cross References

Using Color, 5-8

Topic References

1. Campbell, J. L., Richard, C.M., Brown, J. L., & McCallum, M. (2007). Crash warning system interfaces: Human factors insights and lessons

learned, final report (Report No. DOT HS 810 697). Washington, DC: National Highway Traffic Safety Administration. Available at

www.nhtsa.gov/DOT/NHTSA/NRD/Multimedia/PDFs/Crash%20Avoidance/2007/CWS_HF_Insights_Task_5_Final_Rpt.pdf

2. International Organization for Standardization (1984). Development and principles for application of public information symbols. (ISO/TR

7239). Geneva: Author.

3. ISO (1988). General principles for the creation of graphical symbols, Part I: Graphical symbols for use on equipment (ISO ISO 3461-1).

Geneva: Author.

4. ISO (2009). Road vehicles—Ergonomic aspects of transport information and control systems—Specifications and test procedures for in-

vehicle visual presentation (ISO 15008). Geneva: Author.

5. O’Day, S., & Tijerina, L. (2011). Legibility: Back to the basics. SAE International Journal of Passenger Cars—Mechanical Systems, 4(1),

591-604.

6. Human Factors and Ergonomics Society (2007). Human factors engineering of computer workstations (ANSI/HFES 100-2007). Santa

Monica, CA: Author.

7. Mourant, R. R. & Langolf, G. D. (1976). Luminance specifications for automobile instrument panels. Human Factors, 18(1), 71-84.

8. Howell, W. C., & Kraft, C. L. (1959). Size, blur, and contrast as variables affecting the legibility of alphanumeric symbols on radar-type

displays (WADC Technical Report 59-536). Wright-Patterson Air Force Base, OH: Wright Air Development Center.

9. Giddings, B. J. (1972). Alpha-numerics for raster displays. Ergonomics, 15(1), 65-72.

10. Carney, C., Campbell, J. L., & Mitchell, E. A. (1998). In-vehicle display icons and other information elements: Literature review (Report No.

FHWA-RD-98-164). Washington, DC: Federal Highway Administration.

Temporal Characteristics of Visual Displays

Introduction

The temporal characteristics of visual messages can involve the use of flashing, blinking and motion in order to draw

attention toward a particular visual display. The use of temporal characteristics, such as flash and motion, takes

advantage of features of the human visual system that are especially sensitive to these features. This topic covers the

design of in-vehicle messages that use temporal and movement features.

Design Goal (L2, L3): Use changes in the temporal characteristics of visual displays, such as flashing, blinking

or apparent motion, to command visual attention.

Supporting Design Guidance

The best available research on this topic suggests that this design goal can be met when:

Flashing is used for important, suddenly-occurring, situations (optimal rate is 3-4 times/s).

Multiple flash mode is used for more urgent situations (this mode uses rapid pulses of flash for each

flash cycle).

Flash rate and duty cycle (on-off period) are adjusted as needed to ensure driver comprehension.

Sequential illumination is used to convey motion and/or direction, but keep text stationary.

Other motion cues such as bouncing or zooming are discouraged while in manual driving or during

transitions to or from automated driving because they may unnecessarily increase driver eyes-off-

road time.

Examples of flash rate and duty cycle.

Flash Rate (Frequency): The number of

flashes per second. Example (a) shows lower

flash rate (one flash per second), and

example (b) shows higher flash rate (four

flashes per second).

A high flash rate can be used to convey high

urgency.

Duty Cycle: The percent of time within a cycle

that the sign is in the “on” state. The example

shows a signal with 80% duty cycle (“on” for

0.8s and “off” for 0.2s).

A higher duty cycle can be used for presenting

icons with accompanying text in order to

provide sufficient time for the driver to read the

text.

Complex Flash: Presentation of multiple

flashes with varying “on” and “off” times. The

example shows two one-second bursts with

four pulses per burst. Each burst is separated by

one second of “off” time.

Complex flashes can be used to further increase

perception of urgency.

0.5 1

1.5 2 2.5

Off

Time (seconds)

b) High Frequency

0.5

1.5

2.5

Time (seconds)

a) Low Frequency

Off

80% Duty Cycle

Off

0 1 2 3 4 5

Time (seconds)

0.8s

Off

Time (seconds)

0 1 2 3 4

Discussion

Use of flashing signals is a standard practice in warning system design (ISO, 2005). Much of this basic guidance may

be useful for the construction of DVIs for L2 and L3 automated driving systems. The basic parameters of a flashing

warning are the frequency, contrast, and duty cycle. Frequency refers to the number of times per second the signal

flashes; a considerable amount of human factors research suggests that for conveying urgency, an optimal flash rate is

3-4 times per second (ISO, 2005; Chan & Ng, 2009). It is possible also to modulate each flash within a cycle via a

multiple “fast flash” mode—this has been found to further increase perception of urgency (Chan & Ng, 2009).

Contrast refers to the change in illumination between the “on” and “off” portions of the flash. The duty cycle of the

flash period refers to the relative amount of “on” and “off” time for the flashing signal—this is a relevant parameter if

symbolic and verbal information are conveyed on a flashing warning because drivers require sufficient “on” time to

view the information.

Sequential illumination of display elements can be used to create apparent motion, which can convey directional

information pertinent to warning systems. ISO (2005) provides examples of movement in association with icons to

convey dangerous situations such as emergency vehicles, icy roadways, etc. Motion should only be applied to icons

or symbols; text should be stationary to reduce potential distraction or implied meaning associated with text motion.

Motion cues may affect cognitive load. (Doshi, Cheng, and Trivedi (2009) used a number of motion cues in a HUD,

including bouncing triangular warning signs, zooming warning signs and moving graphical indicators with a bounce

to show excessive speed. The results suggested that motion without contextual information about speed led drivers to

spend more time looking down at the instrument cluster to determine why the alert was being provided than they did

with no alert.

Design Issues

Flashing and motion can be compelling visual display elements and are appropriate for use in higher priority

situations if the message can be provided in appropriate temporal relationship to the message cause so that nuisance

alarms are not perceived. The increasing ease of presenting visual effects such as zooming and bouncing of visual

elements lead to an expanded interpretation of the “flash” concept. Type of motion, however, should not be used as a

code in and of itself; instead, supplementary information should be provided (such as current speed or speed limit),

without requiring the driver to visually refer to other instruments or displays.

Cross References

Multimodal Messages, 4-8; Locating a Visual Display, 5-2; Display Glare, 5-4

Topic References

1. International Organization for Standardization. (2005). Road vehicles—Ergonomic aspects of in-vehicle presentation for transport information

and control systems—Warning systems (ISO/TR 16352). Geneva: Author.

2. Chan, A. H. S., & Ng, A. W. Y. (2009). Perceptions of implied hazard for visual and auditory alerting signals. Safety Science. 47, 346-352.

3. Doshi, A., Cheng, S. U., & Trivedi, M. M. (2009). A novel active heads-up display for driver assistance. IEEE Transactions on Systems, Man,

and Cybernetics—Part B: Cybernetics, 39(1), 85-93.

Chapter 6. Auditory Interfaces

This chapter provides guidance for the design of auditory interfaces. Auditory interfaces are

useful for capturing and directing drivers’ attention and for presenting information to drivers

when they are not attending to a visual display. Consequently, these interfaces are particularly

useful for presenting time-sensitive messages to drivers. Auditory signals can be used to convey

three forms of information (Catchpole, McKeown, & Withington, 2004). First, urgency cues

provide information regarding the criticality of the situation or how quickly drivers need to

respond to the warning. Second, location information identifies where the hazard is located or

where it is coming from. Third, the semantic meaning associated with the signal provides

information to the driver about what is happening or what actions to take in order to avoid a

crash. Even though all three forms of information do not necessarily need to be included in a

single message, each of these components can provide useful information for facilitating rapid

and correct responses.

As drivers of L2, and especially L3, automated driving systems may be engaged in non-driving

related tasks, it cannot be assumed that the driver’s visual attention will be directed towards the

roadway. Auditory signals can be effective for presenting information to the driver because they

can be perceived regardless of the direction of visual attention. These signals can be very

effective in when they are salient, appropriately obtrusive, and their meaning can be understood.

Various characteristics of auditory messages can be modulated to affect the level of salience and

obtrusiveness as well as the perception of urgency, and the type of signal used can facilitate

comprehension of the information being displayed.

There are design tradeoffs, however, that need to be considered in order to avoid unwanted side

effects and ensure the message has its intended effect. Salient, obtrusive sounds can be annoying

if they are presented often, too loudly, or if their characteristics are perceived as annoying. Yet

sounds that are not obtrusive or loud enough may go unnoticed. Also, the meaning of some

auditory signals, such as speech messages and auditory icons (auditory signals that sound like a

real object or event, such as a screeching tire), can be easy to understand, while the meaning of

other types of sounds, such as pure tones, must be learned.

The following topics are included in this chapter.

• Perceived Urgency of Auditory Warnings

• Perceived Annoyance of Auditory Warnings

• Loudness of Auditory Warning Signals

Perceived Urgency of Auditory Warnings

Introduction

This topic provides information for designing auditory warning messages that convey a level of urgency that

matches the urgency of the hazard situation.

Design Goal (L2, L3): Use an auditory warning to clearly communicate a level of urgency consistent with the

urgency of the task, event, or situation.

Supporting Design Guidance

The literature suggests that the attributes listed below may be manipulated in support of the design goal. Note that

this list is not intended to be comprehensive.

To increase the perceived urgency: To decrease the perceived urgency:

Use faster auditory signals (e.g., 6

pulse/s).

Use regular rhythms (all pulses

equally spaced).

Use a greater number of pulse burst

units (e.g., 4 units).

Use auditory signals that speed up.

Use high fundamental frequencies

(e.g., 1000 Hz).

Use random or irregular overtones.

Use a large pitch range (e.g., 9

semitones).

Use a random pitch contour.

Use an atonal musical structure

(random sequence of pulses).

Use more urgent words

(e.g., “Danger”).

Use slower auditory signals (e.g., 1.5

pulse/s).

Use irregular rhythms (pulses not

equally spaced).

Use a fewer number of pulse burst

units (e.g., 1 unit).

Use auditory signals that slow down.

Use low fundamental frequencies

(e.g., 200 Hz).

Use a regular harmonic series.

Use a small pitch range (3 semitones).

Use a down or up pitch contour.

Use a resolved musical structure (from

natural scales).

Use less urgent words

(e.g., “Caution”).

Examples of auditory signals (tones) with urgent (top)

and less urgent (bottom) characteristics.

Many pulses per burst

Long burst duration

Short interburst

interval

Fast pulse rate

Short pulse width

High fundamental

frequency

Short interpulse

interval

Low fundamental

frequency

Few (one) pulses

per burst

Slow pulse/burst rate

Long interburst

interval

Discussion

Varying certain acoustical properties has a strong and consistent effect on a person’s subjective impression of the

urgency of an auditory warning. Accurate portrayal of urgency helps drivers to understand the warning and respond

more effectively. Scaling of the urgency of the warning must be selected with understanding the potential for the

driver to be engaged in a non-driving related task, possibly for an extended period of time prior to warning issuance.

In general, greater perceived urgency of a warning is associated with faster reaction times (Campbell, Richman,

Carney, & Lee, 2004; Suied, Susini, & McAdams, 2008) however, signals that are perceived as more urgent than is

warranted by the situation can result in confusion, distraction, or inappropriate responses, such as overly aggressive

responses. If auditory signals are designed with the proper level of urgency in mind, more effective warnings can be

developed.

Design Issues

Signal attributes that can provide urgency cues include time-varying characteristics, frequency characteristics, and

signal complexity (Campbell, Richman, Carney, & Lee, 2004; Suied, Susini, & McAdams, 2008; Department of

Defense, 2012; Pomerleau et al., 1999; Tan & Lerner, 1995; Marshall, Lee, & Austria, 2007). Some specific

characteristics that affect urgency are pulse rate, fundamental frequency, harmonic content, and (potentially)

intensity. Several studies and guideline documents (e.g., Campbell, Richman, Carney, & Lee, 2004; Suied, Susini, &

McAdams, 2008, (Gonzalez, Lewis, Roberts, Pratt, & Baldwin, 2012, 2007) suggest that increasing the pulse rate

can increase perceived urgency; similarly, increasing the fundamental frequency also increases urgency.

Furthermore, Edworthy, Loxley, & Dennis (1991) found that signals with irregular overtones increased perceived

urgency, while those with regular harmonics decreased urgency. Some studies and guidelines (e.g., MIL-STD-

1472G [Department of Defense, 2012], Pomerleau et al., 1999; Tan & Lerner, 1995) suggest that increasing the

intensity (volume) increases the level of perceived urgency; however, intensity as an urgency cue should be used

with caution. Although intensity can affect perceived urgency, it is not always clearly the case—at least one source

(Lee, McGehee, Brown, & Reyes, 2002). showed that increasing the intensity as a means of presenting higher levels

of urgency did not have a significant effect on the performance of the forward crash warning (FCW). More

importantly, high-intensity auditory signals can be perceived as annoying, which can negatively impact driver

performance as well as acceptability (Wiese & Lee, 2004).

Message semantics can also influence the perceived urgency of an auditory warning. A laboratory study (Guilluame,

Drake, Rivenez, Pellieux, & Chastres, 2002). found that familiar, real alarms used in military aircraft were rated

with different levels of urgency than their synthesized counterparts that had similar acoustic characteristics,

suggesting that the mental representation of the sequence interacts with the acoustic properties in the perception of

urgency. Similarly, the semantic content of speech messages has been shown to interact with loudness in simulated

driving (Baldwin &., & May, 2011). The fewest crashes occurred when drivers received collision warning messages

that included either the low-urgency word “Caution” presented at high intensity or the high-urgency word “Danger”

presented at low intensity, while the most crashes occurred when the word “Danger” was presented at high intensity.

These findings suggest that overall perceived urgency can be elevated without substantially increasing the annoying

effects associated with high-urgency acoustic properties by incorporating high-urgency semantics (whether with

speech or with familiar non-speech signals) into auditory messages that have lower-urgency acoustic characteristics.

Cross References

Perceived Annoyance of Auditory Warnings, 6-4

Topic References

1. Campbell, J. L., Richman, J. B., Carney, C., & Lee, J. D. (2004). In-vehicle display icons and other information elements, Volume I: Guidelines

(Report No. FHWA-RD-03-065). McLean, VA: Federal Highway Administration. Retrieved from

www.fhwa.dot.gov/publications/research/safety/03065/index.cfm

2. Suied, C., Susini, P., & McAdams, S. (2008). Evaluating warning sound urgency with reaction times. Journal of Experimental Psychology: Applied,

14(3), 201-212.

3. Department of Defense. (January 11, 2012). MIL-STD-1472G. Department of Defense Design Criteria Standard: Human Engineering. Washington,

DC: U.S. Government Printing Office. Available at http://everyspec.com/MIL-STD/MIL-STD-1400-1499/download.php?spec=MIL-STD-

1472G.039997.pdf

4. Pomerleau, D., Jochem, T., Thorpe, C., Batavia, P., Pape, D., Hadden, J., . . . Everson, J. (1999). Run-off-road collision avoidance using IVHS

countermeasures (Report No. DOT HS 809 170). Washington, DC: National Highway Traffic Safety Administration.

5. Tan, A. K., & Lerner, N. D. (1995). Multiple attribute evaluation of auditory warning signals for in-vehicle crash avoidance warning systems (Report

No. DOT HS 808 535). Washington, DC: National Highway Traffic Safety Administration. Available at https://rosap.ntl.bts.gov/view/dot/13554

6. Marshall, D. C., Lee, J. D., & Austria, P. A. (2007). Alerts for in-vehicle information systems: Annoyance, urgency, and appropriateness. Human

Factors, 49(1), 145-157.

7. Gonzalez, C., Lewis, B. A., Roberts, D. M., Pratt, S. M., & Baldwin, C. L. (2012). Perceived urgency and annoyance of auditory alerts in a driving

context. Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 56(1), 1684-1687.

8. Edworthy, J., Loxley, S., & Dennis, I. (1991). Improving auditory warning design: Relationship between warning sound parameters and perceived

urgency. Human Factors, 33(2), 205-231.

9. Lee, J. D., McGehee, D. V., Brown, T. L., & Reyes, M. L. (2002). Driver distraction, warning algorithm parameters, and driver response to imminent

rear-end collisions in a high-fidelity driving simulator (Report No. DOT HS 809 448). Washington DC: National Highway Traffic Safety

Administration.

10. Wiese, E. E., & Lee, J. D. (2004). Auditory alerts for in-vehicle information systems: The effects of temporal conflict and sound parameters on driver

attitudes and performance. Ergonomics, 47(9), 965-986. doi: 10.1080/00140130410001686294

11. Guilluame, A., Drake, C., Rivenez, M., Pellieux, L., & Chastres, V. (2002). Perception of urgency and alarm design. Proceedings of the 8th

International Conference on Auditory Display.

12. Baldwin, C. L., & May, J. F. (2011). Loudness interacts with semantics in auditory warnings to impact rear-end collisions. Transportation

Research Part F: Traffic Psychology and Behaviour, 14(1), 36-42.

Perceived Annoyance of Auditory Warnings

Introduction

This topic provides information for designing auditory warning messages that are less likely to annoy drivers yet

still convey an appropriate level of urgency. Urgent sounds often have characteristics that can also be perceived as

annoying. Note that the presentation of other auditory information to the driver (e.g., navigation updates) is not

covered by this topic.

Design Goal (L2, L3): Select auditory warnings that minimally annoy drivers.*

Design Guidance

The best available research on this topic suggests that this design goal can be met when:

The perceived urgency of a sound is matched with the urgency of its referent. Drivers who perceive

the benefits of an obtrusive signal will be less likely to be annoyed by it.

Low annoyance sounds are used for benign situations.

Minimize the rate of false or nuisance alarms to reduce the potential for annoyance.

Repetitious Speech-based warnings are minimally repeated and no more than three times per crash

avoidance situation, and in immediate succession.

Systems use sounds with characteristics that promote perceived urgency more than perceived

annoyance.

*This topic does not apply to auditory warnings that are intended to annoy drivers, e.g., seat belt reminder.

Example of one analytical method for estimating the effect on

perceived urgency and annoyance by varying sound parameters.

The figure on the left illustrates the

relationship between urgency and

annoyance when varying a signal

characteristic, such as frequency, pulse rate,

or volume, of an example auditory warning.

The graph shows linear regressions of

subjective ratings of urgency and annoyance

as described in (Gonzalez, Lewis, Roberts,

Pratt, & Baldwin, 2012).

In this example, the greater slope

of the urgency line indicates that

urgency increases more than

annoyance does when the

parameter is increased.

This graph suggests that, for the

particular auditory warning tested,

the parameter under test should be

increased to convey higher urgency

because it has less impact on

annoyance than on urgency.

Discussion

An important tradeoff exists between alerting and annoying when using auditory warnings. Highly urgent signals

can also be perceived as annoying, and while many sound parameters that increase urgency also increase annoyance,

careful design can create highly urgent sounds that are not overly annoying. The goal is to minimize the annoyance

associated with a warning, balanced by the need to match the urgency of the signal to the urgency of the situation.

This is called “annoyance tradeoff” and should be considered in signal design.

Auditory signals that are perceived to be annoying can increase workload (Wiese & Lee, 2004), be distracting, or

cause the driver to disable the warnings altogether. This problem may potentially be compounded when more than

one safety application, each with its attendant warning, is available in the vehicle. Consequently, designers should

consider the potential for “alarm fatigue” when designing systems with multiple auditory warnings, even when the

individual warnings are designed to minimize annoyance. General Motors Corporation and Delphi-Delco Electronic

Systems (2002) found that forward crash waring (FCW) systems that produce a high number of false alarms can be

considered annoying by drivers, even when the tone is appropriate for a system with a low number of false alarms.

Similarly, although participants in one study (Kiefer, Cassar, Flannagan, Jerome, & Palmer, 2005) considered the

auditory tone to convey the right level of urgency, more than half indicated they would turn off the alert suggesting

that the sound was annoying.

Design Issues

Some sources (Campbell, Richman, Carney, & Lee, 2004; Marshall, Lee, & Austria, 2007; Hellier & Edworthy,

1989) indicate that certain quantifiable sound parameters such as inter-pulse interval (time between pulses), number

of repetitions, duty cycle, and frequency have a greater effect on urgency than on annoyance. Other studies (Wiese

& Lee, 2004; Gonzalez, Lewis, Roberts, Pratt, & Baldwin, 2012), however, found that increasing signal intensity,

frequency, or duty cycle increased annoyance more than urgency. Results from (Gonzalez, Lewis, Roberts, Pratt, &

Baldwin, 2012) suggest that if a signal parameter’s psychophysical relationship with urgency is not stronger than its

relationship with annoyance, it is likely not a viable parameter. The figure on the previous page demonstrates a

method for quantifying the level of annoyance or urgency as a means of determining the relationship between

urgency, annoyance, and the signal characteristics.

While this guidance applies to warnings, many of the concepts may be applied to the design of non-warning

information for the driver of L2 or L3 automated driving systems. For instance, avoiding unnecessary repetition of a

message, and careful selection of sound characteristics can lessen the potential for driver annoyance due to system

messages. However, some uncertainty remains about the specific characteristics that a driver will find annoying in

either the L2 or L3 automated driving environment. Also at issue is the fact that, especially in L3 operations, it will

not be possible to know what the driver is doing. Drivers who are monitoring the road may find an alert annoying,

whereas a driver engaged in some non-driving task may find the same alert helpful.

Topic References

1. Gonzalez, C., Lewis, B. A., Roberts, D. M., Pratt, S. M., & Baldwin, C. L. (2012). Perceived urgency and annoyance of auditory alerts in a

driving context. Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 56(1), 1684-1687.

2. Wiese, E. E., & Lee, J. D. (2004). Auditory alerts for in-vehicle information systems: The effects of temporal conflict and sound parameters

on driver attitudes and performance. Ergonomics, 47(9), 965-986. doi: 10.1080/00140130410001686294

3 General Motors Corporation & Delphi-Delco Electronic Systems. (2002). Automotive collision avoidance system field operation test: Warning

cue implementation summary report (Report No. DOT HS 809 462). Washington, DC: National Highway Traffic Safety Administration.

4. Kiefer, R. J., Cassar, M. T., Flannagan, C. A., Jerome, C. J., & Palmer, M. D. (2005). Surprise braking trials, time-to-collision judgments,

and “first look” maneuvers under realistic rear-end crash scenarios: Forward Collision Warning Requirements Project, Tasks 2 and 3a

Final Report( Report No. DOT HS 809 902). Washington, DC: National Highway Traffic Safety Administration,. Available at

https://one.nhtsa.gov/DOT/NHTSA/NRD/Multimedia/PDFs/Crash%20Avoidance/2005/FCWTask2and3aFinal.pdf

5. Campbell, J. L., Richman, J. B., Carney, C., & Lee, J. D. (2004). In-vehicle display icons and other information elements, Volume I:

Guidelines (Report No. FHWA-RD-03-065). McLean, VA: Federal Highway Administration. Available at

www.fhwa.dot.gov/publications/research/safety/03065/index.cfm

6. Marshall, D. C., Lee, J. D., & Austria, P. A. (2007). Alerts for in-vehicle information systems: Annoyance, urgency, and appropriateness.

Human Factors, 49(1), 145-157.

7. Hellier, E., & Edworthy, J. (1989). Quantifying the perceived urgency of auditory warnings. Canadian Acoustics, 17(4), 3-11.

Loudness of Auditory Warning Signals

Introduction

This topic provides guidance and information regarding the intensity levels for presenting auditory warnings that are

clearly perceivable. In order to be effective, auditory warnings must be loud enough to be heard in the noisy driving

environment. The information below may assist designers in determining appropriate volume levels for presenting

clearly audible warnings to drivers.

Design Goal (L2, L3): Select auditory warning signals that are loud enough to overcome masking sounds from

road noise, the cab environment, and other equipment.

Supporting Design Guidance

The best available research on this topic suggests that this design goal can be met when:

The amplitude of auditory signals is in the range of 10-30 dB above the masked threshold (MT), with

a recommended minimum level of 15 dB above the MT. Alternatively, the signal is at least 15 dB

above the ambient noise.

The signal does not exceed a maximum intensity of 90 dBA.

Designers avoid presenting auditory warnings at more than 30 dBA above the MT to avoid startling

or annoying drivers.

The auditory warning signal includes frequency components in the range of 500-2500 Hz, and the

signal includes at least two dominant components in the subset range of 500-1500 Hz.

The intensity of cautionary warning signals is less than the intensity of the imminent collision

warning signals; however, if doing so will limit the ability of drivers to perceive the cautionary

warning, other signal characteristics could be used to convey lower urgency.

Other sounds produced by the vehicle (e.g., radio or HVAC fans) are muted or disabled while the

warnings are presented, to enhance the audibility of warnings.

Relationship between masked threshold and recommended signal intensity range.

Adapted from Edworthy & Hellier (2000).

This graph shows the frequency domain of a

hypothetical warning signal superimposed

on the MT for noise conditions while

driving

A. Signal limited to 90 dB above the MT

B. Dominant frequency components in

500−2500 Hz range with two in the

500−1500 Hz range

C. Signal has potential to be startling or

annoying because 3 kHz component is

greater than 30 dB above MT

D. Frequency component will likely not

be heard

E. Frequency component may not be

perceived by some

This graph shows a hypothetical scenario for illustrative purposes only. The signal itself is likely to be annoying, and the noise spectrum may

not represent noise in real driving conditions.

100

0 1000 2000 30 00 4000 5000

Power in d

B SPL

Frequency in Hz

Recommended Signal Intensity Range

(10 to 30 dB above Masked Threshold)

Auditory Warning Signal

Masked Threshold

Ideal Minimum Signal Intensity

(15 dB above Masked Threshold)

Discussion

Little published research examining the loudness of warning messages for automated driving systems is available.

The guidance listed here are based on research examining imminent collision warnings (ICWs) and cautionary

collision warnings (CCWs) and should be generally applicable to presenting urgent (e.g., take control now) and

cautionary (e.g., take control soon) messages in L2 and L3 automated driving systems. Additionally, standards on

the maximum presentation limits of ICWs in an automobile (ISO 7731, 2003) are applicable in this setting.

In order for an auditory warning to be clearly perceived, it must be presented at an intensity that is substantially

greater than the MT. The MT represents the minimum intensity level at which a sound presented among masking

“background” noises is audible to a listener. It is important to note that the MT is not necessarily the same as the

ambient noise level, and several factors influence the MT.

Two sources (Lerner, Kotwal, Lyons, & Gardner-Bonneau, 1996); Campbell, Richman, Carney, & Lee, 2004)

recommend the warning intensity should be at least 20 dB and no more than 30 dB above the MT; however, other

sources (ISO, 2005; Lee et al., 2004) indicate that drivers can discern auditory warnings at as little as 10 dB above the

MT, and they recommend that auditory ICWs should be 10

-15 dB above the MT in order for the warning to be

reliably detected. Furthermore, the standards in MIL-STD-1472G (Department of Defense, 2012) require that caution

signals exceed the ambient noise environment by at least 15 dB and that alerting signals exceed ambient noise by at

least 20 dB. An ISO standard (ISO 7731, 2003) regarding danger signals in workplaces requires that at least one of

the following criteria are met in order for non-speech signals to be clearly audible: (1) the A-weighted sound pressure

level (SPL) of the signal must exceed the SPL of the ambient noise by more than 15 dB, (2) the SPL level must

exceed the MT by at least 10 dB in at least one octave band, or (3) the SPL must exceed the MT by at least 13 dB in

at least one 1/3-octave band. Most sources agree that the amplitude of auditory signals for ICWs should not exceed

the MT by more than 30 dB in order to avoid startling or annoying the driver. In any case, the maximum amplitude of

the warning should be limited. For auditory warnings in workplaces, ISO (2005) recommends a limit of 90 dBA.

Design Issues

Meeting these criteria can be challenging in noisy driving environments. If the MT in the vehicle is more than

75 dBA, the warning sound cannot meet the recommended 15 dB above the MT without violating the 90 dBA limit.

One strategy for improving audibility of auditory warnings is to mute in-vehicle systems that generate competing

auditory information or noise (e.g., stereo system or fans) (Kieffer et al., 1999) during warning presentation. Also,

auditory signals comprised of multiple frequencies will increase the likelihood that at least one frequency will be

detected. The ISO standard (ISO 7731, 2003) requires that the signal include frequency components in the range of

500-2500 Hz, and they recommended that there be two dominant components in the range of 500-1500 Hz.

Frequencies in this range fall within the range of hearing that is most sensitive in humans and are most likely to be

detected.

Lee et al. (2004) and Campbell, Bittner, Lloyd, Mitchell, and Everson (1997) recommend that the intensity of

cautionary crash warning (CCW signals should be less than the intensity of ICW signals in order to communicate a

lower level of urgency. Nevertheless, auditory CCWs should follow the same topics for minimum and maximum

intensity as the auditory ICW to ensure that the warning can be detected above other auditory signals in the vehicle

without being annoying or harmful. Signal characteristics other than intensity can be used to convey lower urgency

if lower intensity signals cannot be reliably detected (see 7-4).

Cross References

Perceived Urgency of Auditory Warnings, 6-2

Topic References

1. Edworthy, J., & Hellier, E. (2000). Auditory warnings in noisy environments. Noise Health, 2(6), 27-40.

2. ISO 7731 (2003). Danger signals for public and work areas—auditory danger signals. Geneva: International Organization for

Standardization.

Lerner, N. D., Kotwal, B. M., Lyons, R. D., & Gardner-Bonneau, D. J. (1996, January). Preliminary human factors guidelines for crash avoidance

warning devices (Report No. DOT HS 808 342). Washington, DC: National Highway Traffic Safety Administration. Available at

https://archive.org/stream/preliminaryhuman00lern#page/n1/mode/2up

4. Campbell, J. L., Richman, J. B., Carney, C., & Lee, J. D. (2004). In-vehicle display icons and other information elements, Volume I:

Guidelines (Report No. FHWA-RD-03-065). McLean, VA: Federal Highway Administration. Available at

www.fhwa.dot.gov/publications/research/safety/03065/index.cfm

5. International Organization for Standardization. (2005). Road vehicles—Ergonomic aspects of in-vehicle presentation for transport

information and control systems—Warning systems (ISO/TR 16352). Geneva: Author.

6. Lee, S. E., Knipling, R. R., DeHart, M. C., Perez, M. A., Holbrook, G.T., Brown, S.B., . . . Olson, R.L. (2004). Vehicle-based

countermeasures for signal and stop sign violations: Task 1. Intersection control violation crash analyses. Task 2. Top-level system and

human factors requirements (Report No. DOT HS 809 716). Washington, DC: National Highway Traffic Safety Administration.

7. Department of Defense. (January 11, 2012). MIL-STD-1472G. Department of Defense Design Criteria Standard: Human Engineering. Washington,

DC: U.S. Government Printing Office. Available at http://everyspec.com/MIL-STD/MIL-STD-1400-1499/download.php?spec=MIL-STD-

1472G.039997.pdf

8. Kiefer, R., LeBlanc, D., Palmer, M., Salinger, J., Deering, R., & Shulman, M. (1999). Development and validation of functional definitions

and evaluation procedures for collision warning/avoidance systems (Report No. DOT HT 808 964). Washington, DC: National Highway

Traffic Safety Administration.

9. Campbell, J. L., Bittner, A. C., Jr., Lloyd, M., Mitchell, E., & Everson, J. H. (1997). Driver-vehicle interface (DVI) design guidelines for the

intersection collision avoidance (ICA) system. Seattle, WA: Battelle Human Factors Transportation Center.

Chapter 7. Haptic Interfaces

This chapter contains human factors design guidance on the use of haptic interfaces. There are

two types of haptic interfaces that are discussed in this chapter: vibrotactile and kinesthetic.

Although a full understanding of haptics is not necessary to use the guidance and topics in this

chapter, it will be valuable for users of this document to understand that vibrotactile and

kinesthetic interfaces have fundamental differences that impact how well drivers detect and

understand haptic messages.

Vibrotactile interfaces provide information to the driver using vibrations. Vibrotactile interfaces

need to be in physical contact with the driver to deliver information and, for L2 and L3 vehicles

where the driver is not contacting the pedals or steering wheel during automated driving, likely

work well when included in seat belts or the seats. The term vibrotactile is a combination of two

words, vibration and tactile. The word tactile is used to describe perception of being touched.

Tactile perception is a passive sense as tactile sensations are not necessarily associated with body

movements. This is a defining characteristic between vibrotactile interfaces and kinesthetic

interfaces and it has implications for detectability and understanding. Vibrotactile interfaces are

often used to deliver information that is abstract from haptic signal. In general, people can sense

when (e.g., temporal cue) and where on the body vibrations occur.

Kinesthetic interfaces provide information by causing limb or body motion. An example of this

type of interface is when brake pulse displays cause a sudden jerk motion of the vehicle, causing

the driver’s body to move. The word kinesthetic is used in relation to the ability to sense static

and dynamic body posture (e.g., knowing where your hands are located). Some kinesthetic

display types may help enhance awareness (e.g., vehicle brake pulses alert drivers by causing

entire body motions).

Vibrotactile and kinesthetic interfaces for automated driving systems are discussed throughout

this chapter. The topics addressed in this chapter include:

• Selecting a Haptic Display

• Improving Distinctiveness of Haptic Displays

• Accommodating for Vibrotactile Sensitivity Across the Body

• Generating a Detectable Signal in a Vibrotactile Seat

Selecting a Haptic Display

Introduction

This topic provides information about different types of haptic displays, high-level descriptions, and their potential

uses for safety-critical messages. The additional topics in this chapter provide important information that may be

used when selecting the type of display.

Design Goal (L2, L3): Integrate haptic displays with vehicle controls, seats, motion, or other elements of the

vehicle.

Supporting Design Guidance

The literature suggests that these haptic displays may be used in support of the design goal.

Types, implementations, and potential benefits and drawbacks of different haptic display types.

Haptic Display

Type

Implementation Potential Benefits Potential Drawbacks

Vehicle Brake

Pulse

One or more short

applications of the brakes to

create pulses of deceleration.

Effective, highly detectible,

may result in lower peak

deceleration.

Disruptive; likely to be rated

as annoying.

Vibrotactile

Seat

Vibration provided through

the seat or portion of the

seat.

Large display contact area;

driver likely to be in contact

with display; provides

general alert or temporal cue.

Difficult to provide

directional alerts; complex

alerts require

training/experience.

Seat Belt

Vibration

Vibration provided through

the fabric of a seat belt.

Moderate display contact

area; Most drivers are likely

to be in contact with display;

provides general alert or

temporal cue.

Not every driver wears their

seat belt; difficult to provide

directional alerts; complex

alerts require

training/experience.

Seat Belt

Pre-tensioner

Tightening or tugging of the

seat belt.

Helps orient driver to

forward roadway.

Not every driver wears their

seat belt; may be confused

with other events; alert may

be viewed as disruptive or

annoying.

Note: Vehicle brake pulses and vibrotactile seat displays are most likely to be detected by the majority of drivers. This

is due to the potential for large variability in driver seat positions during automated driving and overall seat belt use

trends in the United States.

Tip for making trade-off decisions between display effectiveness and user acceptance.

Use this heuristic: Intrusive and annoying haptic displays may lead to better response compliance but may reduce

overall user satisfaction.

Note: Vibrotactile haptic displays are often rated to be less annoying and less intrusive compared to auditory displays

and kinesthetic haptic displays like brake pulses and steering wheel torque rotation. Brake pulses more frequently

result in faster reaction times or better selection of appropriate driving maneuvers.

Discussion

Automotive haptic and tactile displays are an emerging area of research and development, especially in regards to

automated driving system driver-vehicle interfaces. The information in this chapter is primarily based on findings

from research examining the use of haptic displays in collision warning systems, international standards for displays,

and expert judgment. This section provides an overview of haptic display types and some of their key

characteristics.

Vehicle brake pulses as a haptic display have been tested in both simulator and on-road studies. One study showed

that brake pulses can be quite effective at getting a driver’s attention and drivers are more likely to detect a brake

pulse if it produces a sensation of “jerk” or “self-motion” (Lee, McGehee, Brown, & Nakamoto, 2012; Brown et al.,

2005). Overall, brake pulses may lead to lower peak deceleration because the vehicle is physically being slowed by

the brake pulses; as a result, drivers may not have to act out hard braking. One usability drawback is that drivers

tend to report that vehicle brake pulses are too disruptive, which can lead to annoyance ratings that are unfavorable.

Vibrotactile seats will be discussed in greater detail in other topics within this chapter. The implication from the

literature is that vibrotactile seat displays may function best as a general alert or temporal cue. Many design issues

limit the use of haptic displays for delivering complex messages; the topic on spatialized vibrotactile seat displays in

this chapter addresses this issue.

The use of seatbelt vibrations as a haptic display may only serve drivers as a temporal cue or general alert. The

correspondence between the warning and where the hazard is located is too tenuous for assuming that locations such

as “forward” are automatically implied.

Seat belt pre-tensioning can prompt the driver to the forward roadway for forward crash warning (FCW) systems

(Forkenbrock et al., 2011). However, the extent that these messages may be confused with other messages or events

must be considered before application in L2 or L3 automated driving settings. Additionally, providing information

through the seat belt must be done with the understanding that seat belt use is only 87 percent in the United States

(Pickrell & Liu, 2014); thus some users will not benefit from this type of display.

Auditory and visual displays can specify the meaning of a message through direct or indirect means. In-vehicle

haptic displays can only deliver messages indirectly, which leaves drivers to interpret the intended meaning. Some

haptic display characteristics can enhance how well drivers interpret the meaning of a haptic signal. There is some

general agreement across the literature that there should be adequate time for drivers to respond if the warning

messages are to be at all more useful than contextual cues from the environment (Lee, McGehee, Brown &

Marshall, 2006; Abe & Richardson, 2006). Adequate timing may help considerably when drivers need to derive a

message from an indirect signal. When haptic signals are in temporal correspondence with a lane departure, general

vibrations from the seat (ISO 17387, 2008) can be quite effective.

Design Issues

Although earlier guidelines generically indicate that tactile warnings may be used to indicate the direction of the

hazard and that this directional cue should be reserved for use during imminent situations (Lerner, Kotwal, Lyons, &

Gardner-Bonneau, 1996), there is no indication that there is an automatic cognitive process that supports this claim.

Studies on tactile warnings provide participants with a large amount of training (Fitch, Hankey, Kleiner, & Dingus,

2011; Jones, Gray, Spense, & Tan, 2008), and experience (Ho, Tan, & Spence, 2005).

Cross References

Selection of Sensory Modality, 4-6; Multimodal Messages, 4-8

Topic References

1. Lee, J. D., McGehee, D. V., Brown T. L., & Marshall, D. (2006). Effects of adaptive cruise control and alert modality on driver performance. Transportation

Research Record: Journal of the Transportation Research Board, 1980, 49-56.

2. Abe, G., & Richardson, J. (2006). The influence of alarm timing on driver response to collision warning systems following system failure. Behaviour &

Information Technology, 25(5), 443-452.

3. Abe, G., & Richardson, J. (2004). The effect of alarm timing on driver behaviour: an investigation of differences in driver trust and response to alarms according

to alarm timing. Transportation Research Part F: Traffic Psychology and Behaviour, 7(4-5), 307-22.

4. ISO 17387 (2008). Intelligent transport systems—Lane change decision aid systems (LCDAS)—Performance requirements and test procedures. Geneva:

International Organization for Standardization.

5. De Groot, S., de Winter, J. C. F., López-García, J. M., Mulder, M., & Wieringa, P. A. (2011). The effect of concurrent bandwidth feedback on learning the lane-

keeping task in a driving simulator. Human Factors, 53(1), 50-62.

6. Fitch, G. M., Hankey, J. M., Kleiner, B. M., & Dingus, T. A. (2011). Driver comprehension of haptic seat alerts intended for use in an integrated collision

avoidance system. Transportation Research Part F: Traffic Psychology and Behaviour, 14, 278-290.

7. Lee, J. D., McGehee, D. V., Brown, T. L., & Nakamoto, J. (2012). Driver sensitivity to brake pulse duration and magnitude. Ergonomics, 50(6), 828-836.

8. Brown, S. B., Lee, S. E., Perez, M. A., Doerzaph, Z. R., Neale, V. L., & Dingus, T. A. (2005). Effects of haptic brake pulse warnings on driver behavior during

an intersection approach. Proceedings of the Human Factors and Ergonomics Society 49th Annual Meeting, 1892-1896.

9. Forkenbrock, G., Snyder, A., Heitz, M., Hoover, R. L., O'Harra, B., Vasko, S., & Smith, L. (2011). A test track protocol for assessing forward collision warning

driver-vehicle interface effectiveness (Report No. DOT HS 811 501). Washington, DC: National Highway Traffic Safety Administration.

10. Pickrell, T. M., & Liu, C. (2014). Seat belt use in 2013: Overall results (Report No. DOT HS 811 875). Washington, DC: National Highway Traffic Safety

Administration.

11. Lerner, N. D., Kotwal, B. M., Lyons, R. D., & Gardner-Bonneau, D. J. (1996). Preliminary human factors guidelines for crash avoidance warning devices.

Interim report (Report No. DOT HS 808 342). Washington, DC: National Highway Traffic Safety Administration.

12. Jones, C. M., Gray, R., Spense, C., & Tan, H. Z. (2008). Directing visual attention with spatially informative and spatially noninformative tactile cues.

Experimental Brain Research, 186(4), 659-669.

13. Ho, C., Tan, H. Z., & Spence, C. (2005). Using spatial vibrotactile cues to direct visual attention in driving scenes. Transportation Research Part F: Traffic

Psychology and Behaviour, 8(6), 397-412.

Improving Distinctiveness of Haptic Displays

Introduction

This topic contains information on how to make haptic displays perceptively distinct. The design goal below

originates from ISO 17387 (2008) that indicates, “…warnings [should be] clearly distinguishable from other signals

of the same type within the vehicle.” This topic describes how designers could comply with the ISO standard when

using haptic displays.

Design Goal (L2, L3): Ensure haptic warnings are clearly distinguishable from other haptic signals in the

vehicle.

Supporting Design Guidance

The best available research on this topic suggests that this design goal can be met when:

Designers select haptic signals of higher intensity than the natural vibrations that also reach the driver

through the vehicle component used to deliver the signal (Ryu, Chun, Choi, & Han, 2010). Note, the

duration of high intensity signals can affect driver comfort levels.

Haptic displays are incorporated in those areas of the vehicle where the driver will remain in contact

with (e.g., seat pan, seat belt).

The duration of high intensity signals is within driver comfort levels.

Multiple simultaneous vibrations across the body are not used (Gallace, Ho, & Spence, 2007).

Apparent motion within vibrational surfaces is used to enhance distinctiveness; apparent motion can

be accomplished by sequentially activating vibrating motors in time-series. Note that for apparent

motion, the minimum distance between tactors motors needs to exceed the two-point threshold (see

Accommodating for Vibrotactile Sensitivity Across the Body, 7-6).

Sufficient training or documentation about function is provided for single systems with multiple

vibrotactile haptic signals used to represent different messages.

Examples of perceptibly different vibrotactile signals in the driver seat

(Fitch, Hankey, Kleiner, & Dingus, 2011).

Warning Type Haptic Signal

Forward Crash Warning (FCW) Two front tactors in the seat-pan simultaneously activated 5 times in a pulse

pattern (200 ms on, 50 ms off pattern).

Curve Speed Warning (CSW) Front tactors simultaneously activate for 1 second.

In-Vehicle Warning (IVW) Two front tactors in the seat-pan simultaneously activated 8 times in a double

pulse pattern (200 ms on, 50 ms off, 300 ms on, 200 ms off).

Lane Change Warning (LCW) Left or right tactor in the backrest activation in a pulse pattern.

Lane Departure Warning (LDW) Left or right tactor in the backrest activate for 1 second.

Note that there is a high degree of training required for drivers to be able to identify multiple unique vibratory

messages.

Discussion

This topic uses examples from the existing literature to illustrate how haptic displays could be designed to comply

with ISO standard 17387 (2008). The design methods mentioned in this topic are not mentioned within the standard.

Drivers may be engaged in a variety of non-driving related tasks while operating in either L2 or L3 automated driving

states (Llaneras, Salinger, & Green, 2013). This may lead to differences in gross body position, in addition to the driver

ceasing to have contact with the vehicle controls (i.e., steering wheel and pedal assembly). Areas such as the seat pan

(see topic 7-6) and seat belt should retain reliable levels of driver physical contact during automated operation.

Research on FCW systems has indicated that seat belt pre-tensioning can prompt the driver to the forward roadway

(Forkenbrock et al., 2011). However, the potential for this message to be confused with other messages or events

must be considered before application in L2 or L3 automated driving settings. Additionally, providing information

through the seat belt must be done with the understanding that seat belt use in the United States is only 87 percent

(Pickrell & Liu, 2014); thus some users will not obtain information from this type of display.

To ensure that vibrational signals from haptic displays are perceptively different from naturally occurring vibrations,

vibrational measurements of the vehicle component that will be used for the haptic display need to be obtained

under natural conditions, and then used to determine the vibrational parameters. In designing vibrotactile displays, it

is important to note that human sensitivity to vibration is highest at frequencies of 200-250 Hz; frequencies above or

below that range require larger amplitude vibrations (Kandel, Schwartz, & Jessell, 2000).

Specific instructions to the driver that indicate there are multiple haptic display information sources within any display

or vehicle may lead to better identification and usage of the signal (Ryu, Chun, Choi, & Han, 2010), but without this

instruction drivers may still benefit from general alerting properties (Fitch, Hankey, Kleiner, & Dingus, 2011).

Apparent motion can be used to cause contrast between the vibrational signals from a haptic display and any natural

vibrations. Natural vibrations do not cause apparent motion for in-vehicle components. Creating apparent motion by

progressively activating tactors within a vibrotactile seat display aids driver responses by enhancing detection (Lee,

McGehee, Brown, & Marshall, 2006).

Design Issues

Vehicle vibrations provide information about vehicle behavior, road conditions, etc. Additional haptic information

should be presented in a manner that is not masked by normal vehicle vibrations.

During L2 or L3 automated driving system operation, drivers may not be optimally positioned to view visual

displays and may not detect some auditory signals. Haptic displays can provide a method of alerting the driver in

these situations.

Avoid sending simultaneous haptic signals to the driver since humans are poor at processing multiple simultaneous

tactile signals (Gallace, Ho, & Spence, 2007). For example, when simultaneous tactile signals are delivered to the

hands and gluteus, a driver may only be able to pay attention to one of those signals. There are some cases when

multiple vibrating surfaces may be helpful for creating temporal redundancy.

Cross References

Designing Messages for Driver Comprehension, 4-2; Accommodating for Vibrotactile Sensitivity Across the Body,

7-6

Topic References

1. ISO 17387 (2008). Intelligent transport systems—Lane change decision aid systems (LCDAS)—Performance requirements and test procedures.

Geneva: International Organization for Standardization.

2. Ryu, J., Chun, J., Choi, S., & Han, S. H. (April, 2010). Vibrotactile feedback for information delivery in the vehicle. IEEE Transactions on Haptics,

3(2), 138-149.

3. Gallace, A., Ho, T. Z., & Spence, C. (2007). Multisensory numerosity judgments for visual and tactile stimuli. Perception & Psychophysics, 69, 487-

501.

4. Fitch, G. M., Hankey, J. M., Kleiner, B. M., & Dingus, T. A. (2011). Driver comprehension of haptic seat alerts intended for use in an integrated

collision avoidance system. Transportation Research Part F: Traffic Psychology and Behaviour, 14, 278-290.

5. Llaneras, R. E., Salinger, J., & Green, C. A. (2013). Human Factors issues associated with limited ability autonomous driving systems: Drivers'

allocation of visual attention to the forward roadway. Proceedings of the International Driving Symposium on Human Factors in Driver Assessment,

Training, and Vehicle Design, 7, 92-98.

6. Forkenbrock, G., Snyder, A., Heitz, M., Hoover, R. L., O'Harra, B., Vasko, S., & Smith, L. (2011). A test track protocol for assessing forward

collision warning driver-vehicle interface effectiveness (Report No. DOT HS 811 501). Washington, DC: National Highway Traffic Safety

Administration.

7. Pickrell, T. M., & Liu, C. (2014). Seat belt use in 2013: Overall results (Report No. DOT HS 811 875). Washington, DC: National Highway Traffic

Safety Administration.

8. Kandel, E. R., Schwartz, J. H., & Jessell, T. M. (2000). Principles of neural science (4th Ed.). New York: McGraw-Hill Health Professions.

9. Lee, J. D., McGehee, D. V., Brown, T. L., & Marshall, D. (2006). Effects of adaptive cruise control and alert modality on driver performance.

Transportation Research Record: Journal of the Transportation Research Board, 1980, 49-56.

Accommodating for Vibrotactile Sensitivity Across the Body

Introduction

This topic describes how a vibrotactile display corresponds to the body’s sensitivity to vibration. A driver’s ability

to perceive a haptic display is a function of vibration amplitude, frequency, duration, surface size, and the body

location where the vibrating surface of the vibrotactile display contacts the driver. Designers need to keep this in

mind when using vibration to deliver information to drivers.

Design Goal (L2, L3): Select a vibration intensity consistent with the sensitivity of the targeted body location.

Supporting Design Guidance

The best available research on this topic suggests that this design goal can be met when:

A larger vibrating surface area is used; this increases perceived intensity for low sensitivity body

regions.

Optimal vibration frequencies are selected; tactile sensitivity is optimal between 200 and 250 Hz

(Kandel, Schwartz, & Jessell, 2000), and highest between 150 and 300 Hz (Jones & Sarter, 2008).

Vibration intensity is increased or decreased by adjusting the vibration frequency or amplitude, but

not both. Note that only relative settings of amplitude and vibration are possible for some types of

vibrotactile displays (e.g., amplitude and frequency are coupled for most tactors that use an eccentric

rotating mass).

Proper tactor placement is used. For vibrotactile seats, use the ratio of the area of the seat where the

message will be presented to the two-point separation threshold between tactor motors to determine

the minimum density of vibrating motors within the vibrating surface area (e.g., Reiner, 2003).

Examples of locations for varying vibrotactile intensity.

Vibration sensitivity across body locations Minimum density of vibrating motors

This graphic is a generalization of vibration sensitivity.

Specifics are left out because of variability due to the

following:

• Individual differences in body mass will impact how

well vibrations are perceived.

• Clothing (e.g., shoes) will also impede vibrations and

affect how well vibrations are perceived.

• Drivers in L2 and higher levels of automation will not

necessarily have contact with the wheel or pedals.

The ratio between seat size and two-point threshold for

the area of the seat picture above results in a 6x4

matrix of tactor motors using the following:

  ( )

   

(

󰂽 

)

= ~ 

  ( )

   

(

󰂽 

)

= ~ 

Least Sensitive

1 - back

1 - gluteus

4 - mid-foot

3 - heel

2 - abdominal area

4 - hand

Most Sensitive

1 2 3 4

Discussion

The information in this topic was assembled using information from physiological research and a synthesis of

research on automotive vibrotactile displays. Basic physiology indicates that mechanoreceptors are distributed

differentially throughout the body, and skin density is not the same across the body. Both these factors influence

how vibrations are felt. Although there are several examples of in-vehicle vibrotactile displays that contact various

body sites (e.g., hands, feet, back, gluteus), empirical research is limited in regards to explaining how to form the

correspondence between vibrations from a vibrotactile display with the sensitivity of body sites. There is general

agreement from both basic and applied research that indicates detection performance improves when vibration

intensity from a vibrotactile display corresponds with physiological sensitivity to vibration.

Basic research has shown that increasing the vibrational surface area of a vibrotactile display increases perceived

intensity (Jones & Sarter, 2008; Cholewiak & Craig, 1984). One way this can be accomplished is by activating more

tactor motors within a larger array of tactors. There is a linear relationship between the number of tactors used to

generate a vibrating surface and a driver’s ability to detect the vibrating surface. More active tactors leads to higher

perceived intensity. This may be a result of the vibration reaching more of the mechanoreceptors within the skin.

When frequency is kept constant but amplitude gets increased, drivers perceive the frequency of the signal to

increase. This perceptual phenomenon is why there are large individual differences across people for their sensitivity

to vibration (Jones & Sarter, 2008). Amplitude and frequency of vibration can be used differentially within certain

vibrotactile displays (e.g., Rosario et al., 2010) but not all (e.g., Ji, Lee, & Hwang, 2011). When testing vibration

parameters, it will be important to note the vibrational elements that accomplish end-user performance goals.

Reiner (2003) calculated the tactor density for a vibrotactile display using the ratio of the size of the driver seat to a

two-point discrimination threshold distance for the driver’s back (e.g., 4 cm). It is advisable that designers measure

the two-point threshold using the seat that will contain the vibrotactile display. This is advised because the two-point

threshold will depend on characteristics that dampen and diffuse the vibration (e.g., seat fabric, cushioning).

Design Issues

There are some additional considerations when selecting the body site that will receive the information. Large body

areas like the gluteus, back and abdominal regions are not often used by people to pick-up information from their

environment (Gallace, Tan, & Spence, 2007). Although the research is limited on how this effects vibrotactile

display types, other haptic displays that deliver messages to the driver through the vehicle control elements

(e.g., steering wheel; Suzuki & Jansson, 2003) tend to enhance response time as a result of better correspondence

between the warning and the required maneuver.

In addition, the correlated effect of frequency and amplitude in perceiving vibration is different across the body.

Displacement has a stronger influence at some locations (e.g., the abdomen) and frequency has a stronger influence

at other locations (e.g., the fingertips) (Jones & Sarter, 2008). This will become a design consideration when

determining whether or not to increase intensity by changing vibration frequency or amplitude, but relevant research

on this consideration is very limited. To support display redundancy, use large vibrating surfaces to ensure the haptic

display makes contact with the driver; e.g., this can be accomplished using multiple vibrating motors embedded in

the seat (Fitch, Hankey, Kleiner, & Dingus, 2011).

Cross References

Selection of Sensory Modality, 4-6

Topic References

1. Kandel, E. R., Schwartz, J. H., & Jessell, T. M. (2000). Principles of neural science (4th Ed.). New York: McGraw-Hill Health Professions.

2. Jones, L. A., & Sarter, N. B. (2008). Tactile displays: Guidance for their design and application. Human Factors, 50(1), 90-111.

3. Reiner, A. (2010). Sensor-actuator supported implicit interaction in driver assistance systems. Wiesbaden, Germany: Vieweg+Teubner.

4. Cholewiak, R. W., &. & Craig, J. C. (1984). Vibrotactile pattern recognition and discrimination at several body sites. Perception and

Psychophysics, 35(6), 503-514.

5. Rosario, H., Louredo, M., Diaz, I., Soler, A., Gil, J. J., Solaz, J. S., & Jordi, J. (2010). Efficacy and feeling of a vibrotactile frontal collision

warning implemented in a haptic pedal. Transportation Research Part F: Traffic Psychology and Behaviour, 13(2), 80-91.

6. Ji, Y. J., Lee, L., & Hwang, W. (2011). Haptic perception in the vehicle seat. Human Factors and Ergonomics in Manufacturing and Service

Industries, 21(3), 305-325.

7. Gallace, A., Tan, H. Z., & Spence, C. (2007). The body surface as a communication system: The state of the art after 50 years. Presence, 16

(6), 655-676.

8. Suzuki, K., & Jansson, H. (2003) An analysis of driver’s steering behavior during auditory or haptic warnings for designing of lane departure

warning system. JSAE Review, 24, 63-70.

9. Fitch, G. M., Hankey, J. M., Kleiner, B. M., & Dingus, T. A. (2011). Driver comprehension of haptic seat alerts intended for use in an

integrated collision avoidance system. Transportation Research Part F: Traffic Psychology and Behaviour, 14, 278-290.

Generating a Detectable Signal in a Vibrotactile Seat

Introduction

This topic contains information on where to place vibrating motors to create a vibrotactile haptic display within the

driver seat. Vibrotactile seat displays appear across the literature for a wide variety of applications. A critical design

element is to ensure the vibration signal is detectible across drivers and the situations they encounter.

Design Goal (L2, L3): Ensure drivers can feel the vibrations from the vibrotactile seat.

Supporting Design Guidance

The best available research on this topic suggests that this design goal can be met when:

There is contact between the driver and the tactile display. To do this, place tactors where seated

pressures are the highest. Use measurements of pressure for an average driver in the seated position

for the seat that will contain the vibrotactile display. Seat pan pressure distribution plots may be

sufficient.

Vibrations are measured at the surface of the seat where the intended message is to occur. This will

help to ensure that the selected intensity surpasses any attenuation accounted by seat materials.

The selected combination of frequency and amplitude has high detectability.

A diverse sample of people is used to test different vibration settings.

Illustration of seat pressure distribution for selecting vibrating surfaces.

The left image depicts a hypothetical seat

pressure distribution and the right image

depicts where tactor motors could be

placed to ensure contact with a driver in

this seat.

Note on the illustration: the pressure distribution shown in

the left image is an artistic rendition. It does not reflect

real data. The tactors in the right image are not to scale.

Examples of the vibration frequency and amplitude for vibrotactile display types at two seat regions.

Detectable vibration measurements from seat pan and and back rest surfaces of a

vibrotactile seat display [1]:

Seat

Region

Frequency

Range

Amplitude

Range

Seat Pan 26 to 30 Hz 2.02 to 2.65 g

Back Rest 30 to 34 Hz 2.65 to 3.38 g

Discussion

The practice of using seat pressure distributions to place vibratory signals at locations of greater seat pressure is

common (Ji, Lee, & Hwang, 2011; Fitch, Kiefer, Hankey, & Kleiner, 2007; Hogema, Sjoerd, Van Erp, & Kiefer,

2009; Reiner, 2010). Three potential reasons for designers and researchers to do this are:

1) Seat-pressure distribution provides an indication of where on the seat the driver is likely to be seated

(Reiner, 2010),

2) The transfer of vibrational energy from the vibrating surface to the receptors in the skin that sense vibration

is more efficient when friction is highest between the vibrating surface and the driver (Dobbins &

McKinley, 2008), and

3) The weight of the driver on the vibrating motor will change the vibration frequency it supplies; this has

implications for the amount of frequency and amplitude that should be used to overcome variability in

driver size and weight and retain a detectable signal. The response of tactors to this loading effect is not

uniform across all tactor types.

Although there are only a few studies where the researchers used an accelerometer to measure vibration at the

surface of the seat (Ji, Lee, & Hwang, 2011; Hogema, Sjoerd, Van Erp, & Kiefer, 2009; Higuchi &

Raksincharoensak, 2010), this is a far better practice than relying on the operating frequency reported by

manufacturers of vibrating tactor motors. Vibrational energy from an embedded tactor has to travel through seat

materials to reach the driver. The attenuation of this energy will correspond with the amount of and type of material

that the vibration has to travel through to reach the driver. Any adjacent material that absorbs the vibrational energy

will cause a similar problem. Thus, any seat material that is positioned between and around the vibrating motor

should be of an appropriate impedance to ensure adequate propagation of the vibrational energy (Dobbins &

McKinley, 2008).

Although only a few relevant studies demonstrate that both vibration frequency and amplitude are important for

making an effective vibrotactile display (Rosario et al., 2010; Ji, Lee, & Hwang, 2011), this concept is well

supported and universally accepted.

There are many individual differences that impact how well people detect vibrations. Some known factors are body

composition and attire. These elements need to be considered when selecting test participants during the design test

phases. Skin mass and clothing impede as well as diffuse vibrational energy.

Design Issues

Vibration Frequency and Amplitude are not always Separately Controllable: Many actuators that are used for

vibrotactile seat displays (e.g., a tactor with a spinning eccentric mass controlled by a motor) have only a voltage

input. With these types of vibrating motors, frequency and amplitude can only be measured rather than controlled

independently (Ji, Lee, & Hwang, 2011). Due to their differing actuation methods, different tactor types have

different operating properties. In addition to the peak frequency of the vibrotactile display, the rise time, duty cycle,

and amplitude range should be considered. Also, properties such as the response of the display to loading (e.g., a

person sitting atop the display) and its surrounding material should be carefully considered as some types of

vibrotactile displays rely upon rotation of an eccentric mass to operate and do not respond as well under loaded

conditions as other vibrotactile display types.

Postural changes while driving may affect the detectability of a vibrotactile display: One researcher suggests the

continual variability in posture can be accounted for by using built-in pressure sensors that detect postural changes.

A system can then subsequently adjust the vibrational surface area to correspond to changes in seated pressure. This

ensures the tactors located at the highest seated pressure zone are active and the signal is appropriately intense

(Reiner, 2010). However, particularly in L3 automated driving systems, the driver’s position and posture are less

certain than in other vehicles where the driver is in a typical driving posture. This change in loading upon the seat

pan may affect the efficacy of a vibrotactile display.

The physical distance between the driver and the vibrating motor plays a role in how much of the signal actually

reaches the driver (Dobbins & McKinley, 2008). The greater the distance, the less of the actual signal reaches the

driver.

Cross References

Accommodating for Vibrotactile Sensitivity Across the Body, 7-6

Topic References

1. Ji, Y. J., Lee, L., & Hwang, W. (2011). Haptic perception in the vehicle seat. Human Factors and Ergonomics in Manufacturing and Service

Industries, 21(3), 305-325.

2. Fitch, G. M., Kiefer, R. J., Hankey, J. M., & Kleiner, B. M. (2007). Toward developing an approach for alerting drivers to the direction of a

crash threat. Human Factors, 49(4), 710-720.

3. Hogema, J. H., Sjoerd, C. D. V., Van Erp, J., & Kiefer, R. J. (2009). A tactile seat for direction coding in car driving: Field evaluation. IEEE

Transaction on Haptics, 2(4), 181-188.

4. Reiner, A. (2010). Sensor-actuator supported implicit interaction in driver assistance systems. Wiesbaden, Germany: Vieweg+Teubner.

5. Dobbins, T., & McKinley, A. (2008). Integration issues of tactile displays in military environments. In J. B. F. van Erp & B.P. Self (Eds.),

Tactile displays for orientation, navigation and communication in air, sea and land environments (RTO Technical Report TR-HFM-122).

NATO Science and Technology Organization.

6. Higuchi, M., & Raksincharoensak, P. (2010). Seat vibrotactile warning interface for forward vehicle collision avoidance. Proceedings of the

SICE Annual Conference 2010, 1370-1373.

7. Rosario, H., Louredo, M., Diaz, I., Soler, A., Gil, J. J., Solaz, J. S., & Jordi, J. (2010). Efficacy and feeling of a vibrotactile frontal collision

warning implemented in a haptic pedal. Transportation Research Part F: Traffic Psychology and Behaviour, 13(2), 80-91.

Chapter 8. Driver Inputs

This chapter provides information related to driver inputs to automated driving system control

functions. Automated driving is an emerging topic and, as such, limited information is available

on specific driver input methods. Therefore, the topics included in this chapter were created from

basic human factors information, previously published guidelines, and current interface

standards.

At a high level, automation controls that must be used while in manual driving (such as those

used to activate automation) can be designed with consideration to reducing the frequency and

complexity of driver interactions when the vehicle is moving. This will minimize driver

distraction and eyes-off-road time during the manual-to-automated driving transition. Optimal

systems can support drivers’ quickly and easily finding the control, discerning how it is used,

and performing the operation with minimal error. Also, the optimal system response following

the driver input is timely and intuitive.

It is difficult to definitively guide the design of controls for automated driving system DVI due to

the multitude of ways that the systems could be implemented. The system may be designed for

one or more specific use cases (e.g., highway speed automation, traffic jam assist, or automated

valet parking), could depend on controls that have multiple uses, and could be designed to

interface with a nomadic device. Each of these system implementations has its own specific

constraints that it imposes on the interface controls.

The following topics are included in this chapter.

• General Guidance for Driver-DVI Interactions

• Control Placement

• Voice Recognition Inputs

General Guidance for Driver-DVI Interactions

Introduction

This topic provides specific information about the design of L2 and L3 automation controls to benefit driver safety

and usability. Ideally, the purpose and operation of well-designed controls will be obvious to drivers ((Lerner,

Kotwal, Lyons, & Gardner-Bonneau, 1996). Controls that are easy to understand and operate minimize distraction

from the driving task during manual to automated driving transitions. Poorly designed controls may adversely affect

or impair the operation of primary driving controls. Therefore, designers’ carefully consider of the placement and

operation of controls in relation to other controls and displays is important (Pomerleau et al, 1999; Stevens et al.,

2005). Since this is a broad topic and little published or publically available research specifically addressing L2 and

L3 automation control design is available, this guideline is primarily supported by basic human factors information.

Design Goal (L2, L3): Provide controls allowing for operation with minimal mental effort, eye glances, and

hand/finger movement.

Supporting Design Guidance

The best available research on this topic suggests that this design goal can be met when controls:

Provide feedback is timely and clear visual, tactile, or auditory feedback for control activation (AAM

2006; Bhise, 2011).

Have identifiable labels (symbols or text) that are visible and located close to the control (Bhise,

2011).

Are easily understood and interpretable (Bhise, 2011).

Placement does not adversely affect or interfere with other critical system components or primary

driving controls.

Size provides a sufficient grasp area and space for hand/finger clearance (Bhise, 2011).

Interaction pacing does not require an uninterruptible series of visual-manual interactions (AAM

2006).

Usage does not compromise the driver’s choice to keep at least one hand on the steering wheel at all

times. AAM principles state that controls on the steering wheel should not require simultaneous

inputs from both hands unless one of the hands only requires a single finger input (AAM 2006).

Methods of control evaluation (Bhise, 2011).

1. Reference and apply available methods, tools, models, customer feedback databases, design guidelines,

principles, and standards.

2. Develop and apply ergonomic checklists.

3. Conduct a task analysis of activities involving the control operation, to break down tasks into subtasks, and to

look for areas of improvement or situations where driver errors may occur.

4. Conduct in-vehicle evaluations where drivers perform a set of tasks using the control. Data collected can include:

time to use the control, errors made, driver likes and dislikes, etc.

5. Include other competitor’s products in the above evaluations to provide benchmarking information.

Discussion

Feedback: The feedback or confirmation provided by the system following driver input should be timely and clearly

perceptible (AAM 2006). Feedback could include a response to control activation, such as the physical and auditory

click of a button press, or a response from a system display, such as an informational dialog box or a system status

indicator icon. Timely and perceptible responses allow the driver to quickly determine that the system is reacting as

expected and that the change in the system is in reaction to their input. This allows drivers to understand when they

have successfully activated automated driving without making second (and perhaps conflicting) inputs or having

uncertainty about the system status.

Identification: Controls are often labeled with the control function, its settings, or both. SAE Recommended Practice

J1138 (SAE, 2009) provides information on when control functions or settings should be labeled.

Interpretation: Automation controls should be intuitive to operate. The control design and movement should match

the driver’s expectations of how the system and application function, especially if mapped with similar functions

such as ACC or lane keeping.

Placement: Automation controls should be placed such that the operation of the control does not adversely affect the

operation of a primary driving control, especially during the transition into or out of automated driving. More

information can be found in Control Placement, 8-4.

Size: The controls themselves need to be physically usable by the driver. This includes designing space for the driver

to grasp the control and space around the control to allow the driver to operate it. As L2 and L3 systems can support

mobility across multiple user groups, additional testing with different populations may be required to ensure controls

are adequately sized.

Pacing: AAM (2006) provides a range of guidance regarding the pacing of control interactions; these “principles”

are generally applicable to the design of the automation DVI. Drivers should be able to control the pace of their

interaction with the system and the timing of the system prompts should be predictable for the task operation.

Drivers should have the option to not respond, delay a response, or temporarily suspend system prompts altogether.

The prompts should not convey that a response is needed urgently and that only one response is possible. If drivers

are interrupted, they should be able to resume at the point of interruption or another logical point. If the system times

out after a reasonable length of time, it should default in a predictable and appropriate way.

Usage: Overall, manual adjustment of controls should not interfere with a driver’s ability to drive safely. In many

driving situations, the vehicle can be driven safely with only one hand on the steering wheel, provided the other

hand is immediately available for steering if it becomes necessary. Interactions should be designed to require that

only one hand at a time needs to be removed from the steering wheel (Stevens et al., 2005).

Design Issues

Complex interactions, such as initial control settings, should be reserved for times when the vehicle is stopped. One

way to prevent certain interactions when the vehicle is moving is to use variable-function keys (i.e., keys that are

mapped to more than one function based on context) on a keypad or touch-screen. Since these keys are

programmable, complex control functions can be made available only at appropriate times. The designer must use

good judgment to determine which interactions should be allowed or denied while the vehicle is in motion

(Pomerleau et al., 1999; Campbell, Carney, & Kantowitz, 1998). The system should clearly indicate those functions

that are not for use while driving and when they are inaccessible (AAM 2006).

Cross References

Control Placement, 8-4; Voice Recognition Inputs, 8-6

Topic References

1. Lerner, N. D., Kotwal, B. M., Lyons, R. D., & Gardner-Bonneau, D. J. (1996, January). Preliminary human factors guidelines for crash avoidance

warning devices (Report No. DOT HS 808 342). Washington, DC: National Highway Traffic Safety Administration. Available at

https://archive.org/stream/preliminaryhuman00lern#page/n1/mode/2up

2. Pomerleau, D., Jochem, T., Thorpe, C., Batavia, P., Pape, D., Hadden, J., . . . Everson, J. (1999). Run-off-road collision avoidance using

IVHS countermeasure (Report No. DOT HS 809 170). Washington, DC: National Highway Traffic Safety Administration.

3. Stevens, A., Hallen, H., Pauzie, A., Vezier, B., Gelau, C., Eckstein, L., . . . Höfs, W. (2005). European statement of principles on the design

of human machine interaction (ESoP 2005, Draft). Directorate-General Information Society Technologies and Media of the European

Commission.

4. Alliance of Automobile Manufacturers (AAM). (2006). Statement of principles, criteria and verification procedures on driver interactions

with advanced in-vehicle information and communication systems, including 2006 updated sections [Report of the Driver Focus-Telematics

Working Group]. Retrieved from www.autoalliance.org/index.cfm?objectid=D6819130-B985-11E1-9E4C000C296BA163

5. Bhise, V.D. (2011). Ergonomics in the automotive design process. Boca Raton, FL: CRC Press.

6. SAE. (2009). SAE J1138: Design criteria—Driver hand controls location for passenger cars, multipurpose passenger vehicles, and trucks

(10,000 GVW and under). Warrendale, PA: SAE International.

7. Campbell, J. L., Carney, C., & Kantowitz, B. H. (1998). Human factors design guidelines for advanced traveler information systems (ATIS)

and commercial vehicle operations (CVO) (Report No. FHWA-RD-98-057). Washington, DC: Federal Highway Administration.

Control Placement

Introduction

This topic provides information about the placement of automation-specific controls in the vehicle. The application

and placement of these controls must not interfere with the primary task of driving the vehicle. Drivers are best-

supported when they can find, reach, and comfortably use controls in the location that they are placed. There is a

lack of published research on the placement of automation controls; therefore, this topic primarily reflects more

general in-vehicle control placement information and design convention.

Design Goal (L2, L3): Ensure control placement and operation does not interfere with the driving task or the

use of other driving controls.

Supporting Design Guidance

The best available research on this topic suggests that this design goal can be met when controls:

• Are easy for drivers to reach and find (Bhise, 2011).

• Are located in a visible area or able to be found blindly (Bhise, 2011).

• Do not obstruct the driver’s field-of-view, vehicle displays, or other vehicle controls (AAM, 2006).

• Are placed such that they are not obstructed by other vehicle controls or displays (Bhise, 2011).

• Do not require the driver to reach their whole hand through the steering wheel (AAM, 2006).

The following figure shows the ideal zone for locating controls within the vehicle.

Diagram of ideal control placement region relative to the driver (Bhise, 2011).

Taylor and Francis Group, LLC, a division of Informa plc.

Midpoint of centroids of

left and right eyel lipses

35 Degree down angle line

Straight-ahead si ghtline

(horizontal )

35°

(See Discussion)

Vertical axis

of the cone

Maximum reach envelope

Discussion

Little published research on the placement of L2 and L3 automated driving system specific controls is available.

However, guidance from standards-setting organizations (AAM, 2006; SAE, 2007; ISO, 2005) and current industry

practices with regard to L1 automated systems (specifically, ACC) can provide helpful information for the

placement of automation controls.

Controls should be easy for the driver to reach and operate. As shown in the figure on the previous page, controls

should be located within the driver’s maximum reach envelope. A reach envelope is a depiction of the area that

drivers can reach within the vehicle, usually represented by a sector of a circle drawn in front of the seated driver.

The values for the maximum reach envelope are described in SAE Recommended Practice J287 (SAE, 2007). The

reach distance data in that standard reflects a distance at which drivers can grasp a knob, rather than simply touch a

control.

Controls should ideally be located above the 35-degree down-angle cone, which is constructed by a line 35 degrees

below the horizontal straight-ahead sightline, through the midpoint of the two eyellipse centroids, and rotated around

the vertical axis through that midpoint. A 30- to 35-degree cone is the limit of the area where drivers can look down

for a control and still detect stop lamps of lead vehicles (Bhise, 2011). Stevens et al. (2005) and ISO (2005) agree

that controls that require lengthy interactions should be placed within 30 degrees of the driver’s normal field of

view. Controls outside of that area should be able to be found blindly. In addition, frequently used controls should

be placed within easy reach and in alignment with the forward view in order to reduce glance times.

Automation controls should not obstruct or interfere with the use of other controls or displays. Stevens et al. (2005)

provides good and bad examples of control design. A good design would incorporate controls that are located within

fingertip reach of the steering wheel. In contrast, a poorly designed control might include a rotary control

concentrically mounted on the steering wheel that requires enough activation force to inadvertently induce a change

in steering angle when activated. There is an area of the instrument cluster that can be viewed through the steering

wheel; however, if controls are placed in this area, drivers must be able to operate them without reaching their hand

through the steering wheel.

The implementation of ACC controls provides useful lessons. Common configurations for ACC controls in

production vehicles, including controls for changing headway settings, are found on the outer edge of the steering

wheel. This location allows for drivers to more easily manipulate the controls. Placing frequently used low-priority

controls (e.g., radio station seek controls) directly adjacent to safety-related controls, such as a gap sensitivity

control, is not recommended. The reason for this is that in the course of using the low-priority control, drivers could

inadvertently and unknowingly change the settings of the automation control, which could result in the vehicle

operating differently than how the driver expects.

Design Issues

Nomadic device (i.e., smartphone) vehicle integration and pass-through voice recognition is becoming more

common in new vehicles. When coupled with an L2 or L3 automated driving system, nomadic devices have the

potential to directly interact with automation functions, such as providing a means of inputting a navigation

destination. These devices provide unique challenges for the DVI designer as the devices change much more rapidly

than vehicle software or platforms, and due to the fact that drivers may interact with the devices in a manner

different than how they expect to interact with the vehicle. This is a rapidly developing area and further research is

needed prior to issuing any stronger guidance for the DVI.

Topic References

1. Bhise, V. D. (2011). Ergonomics in the automotive design process. Boca Raton, FL: CRC Press.

2. Alliance of Automobile Manufacturers (AAM). (2006). Statement of principles, criteria and verification procedures on driver interactions

with advanced in-vehicle information and communication systems, including 2006 updated sections [Report of the Driver Focus-Telematics

Working Group]. Retrieved from www.autoalliance.org/index.cfm?objectid=D6819130-B985-11E1-9E4C000C296BA163.

3. SAE. (2007). SAE J287: Driver hand control reach. Warrendale, PA: SAE International.

4. Stevens, A., Hallen, H., Pauzie, A., Vezier, B., Gelau, C., Eckstein, L., . . . Höfs, W. (2005). European statement of principles on the design

of human machine interaction (ESoP 2005, Draft). Directorate-General Information Society Technologies and Media of the European

Commission.

5. International Organization for Standardization. (2005). Road vehicles—Ergonomic aspects of in-vehicle presentation for transport

information and control systems—Warning systems (ISO/TR 16352). Geneva: Author.

6. General Motors Corporation. (2005). Automotive collision avoidance system field operational test (ACAS FOT) final program report (Report

No. DOT HS 809 886). Washington, DC: National Highway Traffic Safety Administration.

Voice Recognition Inputs

Introduction

This topic addresses speech interactions between drivers and in-vehicle systems. It focuses on one-way speech input,

from the driver to the in-vehicle system. Although speech interactions may place cognitive demands on the driver,

unlike visual-manual interactions, they do not require drivers to take their eyes off the road and their hands off the

steering wheel. Some research has indicated that voice recognition systems are associated with increased off-road

glances (Owens, McLaughlin, & Sudweeks, 2011). Additionally, several simulator and on-road studies of speech

interfaces reported better lane-keeping, fewer glances away from the roadway, shorter glance durations, and lower

subjective workload when compared with visual-manual interfaces. However, some research has indicated that voice

recognition systems are associated with increased off-road glances (Lo & Green, 2013) and, therefore, it cannot be

assumed that using a speech input device does not have any consequences for driver performance. Performing

secondary tasks using speech input still utilizes cognitive resources and, therefore, does not eliminate risk

(Lumsden, 2008). The precise effects of specific system parameters and implementations of speech input on driver

workload and driver performance are largely unknown.

Design Goal (L2, L3): Implement speech controlled in-vehicle systems that have minimal input constraints,

provide user feedback, and have an error handling strategy.

Supporting Design Guidance

The best available research on this topic suggests that this design goal can be met when the user-input features

accommodate the following user requirements:

Conversation Style:

− A natural conversation flow and cadence between user and system is accommodated.

− Input vocabulary is minimized and is consistent with a terse interaction style.

System Feedback:

− When starting an interaction, there is a notification that the system is ready for input.

− After the user provides input, there is feedback so the user knows their input was received.

Error Handling:

− Speech input is only used when the consequences of recognition errors are low.

− The accuracy of the speech recognizer can affect driving performance (see table below).

General Design Goals (Lumsden, 2008) to consider:

− Reduce the user’s cognitive load during transitions between automated and manual driving.

− Reduce interaction time.

− Increase task completion rate.

− Use feedback to the user that reinforces correct use.

− Design so that the users can form a mental model of the system behavior.

− Design so that the system is both effective for an experienced user and appealing for a new user.

The accuracy of the speech recognizer, when considered with the number of steps required to complete a task, has a

large effect on the success rate of task completion. Shown in the table below is the probability of completing an

interaction using only as many steps as are required, given a fixed accuracy level for each step.

Number of steps required to complete an interaction

Accuracy level

of each step

0.99

0.98

0.97

0.96

0.95

0.90

0.86

0.81

0.77

0.90

0.81

0.73

0.66

0.59

0.80

0.64

0.51

0.41

0.33

0.70

0.49

0.34

0.24

0.17

0.60

0.36

0.22

0.13

0.08

0.50

0.25

0.13

0.06

0.03

Discussion

The available literature related to in-vehicle speech input is limited in empirical strength. Most of the design guidance

listed above were derived from studies with modest sample sizes, individual interface usability studies, the judgment of

experts in the field, or unpublished research results. Additional research in this area is needed to investigate the long-

term effects of real-world systems with varying accuracy/error rates, to fully determine the behavioral implications of

new speech technologies.

Conversation style: Driver interactions with the system should be paced by the driver (Lumsden, 2008). This leaves the

driver free to prioritize safe operation of the vehicle over responses to the in-vehicle system. When speaking, users

should be able to converse with the system using a normal conversational cadence. Drivers should not be required to

exaggerate their speech or insert artificial delays between words (Department of Defense, 2012). MIL-STD-1472G

(Department of Defense, 2012) states that “input vocabulary shall be minimized, consistent with system needs, and

selected to provide phonetically distinct elements to eliminate misinterpretation.” Additionally, in a study of user

preferences and responses to scenarios requiring speech input, 93 percent of commands issued were terse (Aldridge &

Lansdown, 1999).

System feedback: Lumsden (2008) recommends that the interaction is initiated with a push-to-talk (PTT; press and

release) button followed by a listening tone. This teaches the user to wait for the tone to begin speaking, which is

beneficial for the recognition accuracy of the speech recognizer. Manual system input has clear proprioceptive, visual,

and manual feedback; however, the feedback provided after giving speech input is less apparent. The system should

provide feedback to the user so they know that their input has been received and understood (Department of Defense,

2012). Popp and Färber (1993) investigated methods of providing feedback for non-transparent driver control actions

(i.e., actions that have long delays before reactions or barely noticeable reactions) performed using speech input. When

providing feedback to nontransparent driver speech input in a simulator, they found that an independent signal tone

without visual feedback produced the least negative effects. The visual display without signal tone also performed well.

These actions both relate to adjustments in L2 and L3 automated driving systems when a system adjustment may not be

immediately apparent.

Error handling: MIL-STD-1472G (Department of Defense, 2012) provides that speech recognition systems may be

used when “the consequences of recognition errors are low” and “identifying and correcting errors would be easy.” In a

simulator study, rejection errors (where the system did not respond to input) caused a smaller proportion of overall task

errors (5%) and were more robust against changes in system accuracy levels than substitution errors (where the system

matches the user input to an incorrect vocabulary word) (Gellatly & Dingus, 1998) Additionally, systems should

require little user training (Department of Defense, 2012).

General design goals: Lumsden (2008) provides a list of general design goals for automotive speech interfaces. It

should be noted that although the guidance is consistent with that found in other sources, the list is made from the

experience of the authors and their unpublished research. Drivers may still have off-road glances while providing voice

input. An examination of a production-type (modified) automotive voice recognition system for text messaging found

that voice interactions increased non-driving related glances and increased driver mental workload (Owens,

McLaughlin, & Sudweeks, 2011).

Design Issues

A simulator study of an in-vehicle speech system (Kun, Paek, & Medenica, 2007), found that the recognition accuracy

of the system affected lane position, when the accuracy level was very low (44%). Push-to-talk (PTT) button use also

affected driving performance when the recognition rate was low. When testing recognition rates of 90 percent, 75

percent, and 60 percent in a simulator, Gellatly and Dingus (1998) found that recognition rate did not affect driving

performance except for the lowest rate (60%). When looking at studies of recognition accuracy, it appears that the

systems function well even with relatively low accuracy levels; however, some negative effects on driving performance

have been reported. These issues are primarily relevant during transitions between different automation states, however

they should be considered during the design of the DVI.

Topic References

1. Owens, J. M., McLaughlin, S. B., & Sudweeks, J. (2011). Driver performance while text messaging using handheld and in-vehicle systems.

Accident Analysis & Prevention 43, 939-947. doi: doi:10.1016/j.aap.2010.11.019

2. Lo, V. E-W., & Green, P. A. (2013). Development and evaluation of automotive speech interfaces: Useful information from the human

factors and related literature. International Journal of Vehicular Technology, 2013, Article ID 924170. Retrieved from

www.hindawi.com/journals/ijvt/2013/924170/

3. Lumsden, J. (2008). Handbook of research on user interface design and evaluation for mobile technology, Volume I. Retrieved from

www.comm.rwth-aachen.de/files/lumsden122.pdf

Department of Defense. (January 11, 2012). MIL-STD-1472G. Department of Defense Design Criteria Standard: Human Engineering. Washington,

DC: U.S. Government Printing Office. Available at http://everyspec.com/MIL-STD/MIL-STD-1400-1499/download.php?spec=MIL-STD-

1472G.039997.pdf

5. Aldridge, L. C., & Lansdown, T. C. (1999). Driver preferences for speech based interaction with in-vehicle systems. Proceedings of the

Human Factors and Ergonomics Society Annual Meeting, 977-981.

6. Popp, M.M. & Färber, B. (1993). Feedback modality for nontransparent driver control actions: Why not visually? In A. G. Gale, I. D.,

Brown, C. M. Haslegrave, H. W. Kryusse, & S. P. Taylor (Eds.), Vision in Vehicles—IV (pp.263-270). Amsterdam: Elsevier Science.

7. Gellatly, A. W., & Dingus, T. A. (1998). Speech recognition and automotive applications: Using speech to perform in-vehicle tasks.

Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 1247-1251.

8. Kun, A., Paek, T., Medenica, Z. (2007). The effect of speech interface accuracy on driving performance. Proceedings of the 8th Annual

Conference of the International Speech Communication Association (Interspeech 2007), 2332-2335.

Chapter 9. Glossary

Adaptive Automation

A system operating under an adaptive function allocation system which reallocates

functions based on the combination of the operating environment and operator loads and

performance.

Adaptive Function Allocation

An automation strategy building on the complementary approach, which assumes criteria

to determine whether functions have to be reallocated, how and when based on changes

in the operating environment, loads or demands to operators and performance of

operators (Inagaki, 2003). An automation system that operates under an adaptive function

allocation is called adaptive automation.

Adaptive Cruise Control (ACC)

A speed control system that automatically adjusts vehicle speed to maintain a set

headway to vehicles ahead without the driver having to use the accelerator or brake.

When the sole automation in function (a lane centering system is not engaged), this is a

form of Level 1 automation.

Apparent Motion

A sensation of motion that is associated with the serial activation of static vibrating

surfaces. The serial activation of adjacent vibrations causes a sensation that the vibrating

surface is moving when, in fact, it is not.

Arcminute

One-sixtieth (1/60) of one degree (1°).

Auditory Icon

An auditory signal that sounds like a real object or event, such as a car horn.

Automation

A device or system that accomplishes all or part of a function that was previously, or

could be, performed, in part or fully, by a human operator (Parasuraman, Sheridan, &

Wickens, 2000).

Automation Etiquette

“The set of prescribed and proscribed behaviors that permits meaning and intent to be

ascribed to the actions [of the automation]” (Parasuraman & Miller, 2004).

A-weighted Sound Pressure Level (SPL)

A measure of the intensity of sound using decibels (dB) referenced to sound pressure

level (SPL). The measurement is weighted across the frequency spectrum using a profile

that accounts for the relative loudness perceived by human hearing.

Calibrated Trust

A state where the automation user’s trust in the automation, as well as their use of the

automation, is appropriately adjusted to the actual performance of the automation

(McGuirl & Sarter, 2006).

Clumsy Automation

A design or situation in which an easy task is automated while a more difficult task is left

for human performance, without assistance (Wiener, 1989).

Compensatory Principle

An automation strategy in which assigning system tasks is based on inherent capability so

that automated tasks are those that machines are better at than humans while manual tasks

are those that humans perform better machines (Fitts List; Fitts, 1951).

Complex Flash

In a visual display, the presentation of multiple signal flashes with varying “on” and “off”

times.

Complexity Creep

Refers to the practice of adding features and capabilities over time to systems, potentially

reducing the interpretability of information, reducing the predictability of the system, and

and/or slowing response time.

Comprehension

The perceptual and cognitive processes by which drivers interpret the meaning of a DVI

message. See also Message Comprehension.

Driver Distraction

A diversion or competing activity which takes the driver’s attention away from activities

which are critical for safe driving.

Driver-Vehicle Interface (DVI)

The means by which a driver gains information about the state of the vehicle (through

auditory, haptic/tactile, and visual displays) and provides control inputs to the vehicle.

Driver Workload

A psychological concept that represents the proportion or amount of mental and physical

capacity required by the driver to complete a task. Workload encompasses both driving

task and situational factors.

Driving Task

A sequence of actions taken by a driver leading to a goal; the driver will normally persist

in these actions until the goal is reached (AAM, 2006).

Duty Cycle

In a visual display, the percent of time within a cycle that the alert signal is in the “on”

state.

Dynamic Function Allocation

An non-static automation strategy which is complementary, enabling the human and

automation to trade-off which functions each performs based on the current situation.

Etiquette

A training approach using a set of heuristics that may enhance and improve a driver’s

interaction with the system, possibly leading to higher levels of both acceptance and

performance.

False Alarm

An alarm that indicates a threat is present when, in fact, no threat exists

Flash Rate (Frequency)

In a visual display, the number of signal flashes per second.

Fundamental Frequency

The lowest frequency in a periodic signal.

Glare

A visual phenomena which occurs when the visual adaptation level is substantially less

than the intensity of a light source within the visual field, causing physical discomfort or

pain (discomfort glare) and/or reduced visibility (disability glare).

Harmonic Content

The harmonics contained in a complex waveform. Complex waveforms consist of a

fundamental frequency and one or more frequencies greater than the fundamental.

Harmonics are frequencies that are integer multiples of the fundamental frequency.

Harmonic Series

The sequence of pitches derived from the multiples of its fundamental, or lowest,

frequency.

Head-Up Display (HUD)

A visual display that presents data in the area of the windshield in the driver’s direct

forward view. An image, such as a collision warning icon, is projected onto a transparent

surface—usually an angled, flat piece of glass or even the windshield itself—which

reflects the image toward the driver. Optically, the image is usually collimated to produce

an image that is perceived to be at infinity. This configuration allows the driver to see and

focus on the display without looking away from their normal viewpoint.

High Head-Down Display (HHDD)

A type of visual display that is mounted below the driver’s normal line of sight (i.e.,

below the horizon line) but within near peripheral vision. These displays are typically

mounted on the dashboard.

Human Factors

An applied, scientific field of study to understand the relationship between devices and

systems and their users, with the capabilities and limitations of human beings as the

central focus.

Kinesthetic Interface

An interface that provides information to the driver by causing limb or body motion such

as when the accelerator pedal “pushes back” against the driver’s foot.

Lane Centering Systems (LCS)

A driver assistive system that identifies the lane center and provides a steering torque to

maintain the vehicle path along the lane center. When the sole automation in function

(ACC is not engaged), this is a form of Level 1 automation.

Lane Keeping Systems (LKS)

A driver assistive system that detects lane markings and provides a steering torque to

prevent the vehicle from crossing a lane marking.

Lateral Warning Systems

Collision warning systems, such as blind spot warnings (BSW) or lane change merge

warnings (LCM), that provide information about hazards in the adjacent lanes or hazards

approaching from the side. These warning systems are primarily dependent on lateral

motion and acceleration to determine when to present warnings.

Legibility

Legibility goes beyond visibility or detection; it implies being able to discern shape or

character identity based on appearance.

Level of Automation

A taxonomy providing for definitions of different functional levels of automation.

NHTSA has defined 5 levels of automation, ranging from no automation (Level 0) to

driver-optional automation (Level 4). As the level of automation increases, a greater

number of driving subtasks are automated. Level 2 automation provides hands-off/feet-

off automation but requires the driver to monitor performance. Level 3 automation

provides hands-off/feet-off automation and does not require the driver to monitor

performance.

Longitudinal Warning Systems

Collision warning systems, such as forward collision warnings (FCW), that provide

information about hazards directly ahead or behind the vehicle. These warning systems

are primarily dependent on longitudinal motion and acceleration to determine when to

present warnings.

Low Head-Down Display (LHDD)

A type of visual display that is mounted further below the driver’s normal line of sight

(i.e., below the horizon line) than is found in a HHDD. These displays are typically

mounted in places such as the instrument panel or the center console, which places the

displayed content farther in drivers’ peripheral vision than a HHDD or HUD.

Luminance

The luminous intensity per unit area of light measured as candela per square meter

(cd/m

Masked Threshold (MT)

The quietest level of a signal that can be perceived in the presence of noise.

Mental Model

An individual’s thought processes and representation about how the automated system

works in the real world.

Message Complexity

Refers to the quantity and variety of basic information elements contained within a

message, as well as the relationships between these elements, how the driver will use the

information, and the value of the information.

Message Comprehension

Refers to the perceptual and cognitive processes by which drivers interpret the meaning

of a message presented through a DVI and includes three stages: extractions, recognition,

and interpretation.

Message Priority

The order of presentation of two or more in-vehicle messages with the order indicating

level of importance.

Mode Awareness

The state of the user correctly understanding under what automation state the system is

operating.

Mode Error

An error resulting from a failure of mode awareness. When a user forgets what state of

automation the system is in and performs an action appropriate to a different state.

Motion Cue

Visual display alert which contains an element of motion within the “on” state, including

bouncing, zooming, and graphical movement.

Multimodal Message

A message consisting of more than one type of signal from the visual, haptic, and

auditory modalities.

Nuisance Alarm

An alarm that correctly indicates a threat is present but the driver does not believe the

alarm is warranted or needed.

Occlusion

Obstructing the view of an interface, driving scene, or other area of interest.

Out-of-the-loop Phenomenon

A situation where the driver is removed from the normal control process due to

automation and experiences a subsequent reduction in his or her awareness of the

vehicle’s current state and level of performance (Kaber & Endsley, 1997).

Overtone

Complex waveforms consist of a fundamental frequency and one or more frequencies

greater than the fundamental. Overtones are any frequencies that are greater than the

fundamental. Harmonics, which are frequencies that are integer multiples of the

fundamental, are a special type of overtone.

Psychophysical Relationship

The relationship between a stimulus and the perceptions and sensations induced by the

stimulus.

Reach Envelope

The area within the vehicle that drivers can reach, usually represented by a sector of a

circle drawn in front of the seated driver. The values for the maximum reach envelope are

described in SAE Standard J287 and reflect the distance at which drivers can grasp a

knob, rather than simply touch a control.

Redundancy

More than one cue for a specific threat or hazard which can be given at the same time

using two or more modes or repeated across time with either the same or different modes.

Residual Function Principle

The earliest automation philosophy in which the automation is designed to perform as

many functions as possible while the remaining functions are performed by the human.

Resolved Musical Structure

A sequence based on natural scales, which consist of musical notes ordered by pitch (e.g.,

tones with frequencies that correspond to notes that can be played on a piano).

Seat Pressure Distribution

The pressure from a seated person acting as a force on a seat (e.g., how the pressures

from body weight are distributed across a seat-pan and back-rest).

Secondary Task

A voluntary task which distracts a driver’s attention away from the main driving task at

hand.

Semantic Content

The meaning contained in an auditory message, whether presented as a speech message

or other type of auditory signal.

Semitone

The smallest musical interval that is commonly used in Western music. Mathematically,

the frequency of a semitone is 2

1/12

times the frequency of the fundamental when the

semitone is a higher pitch than the fundamental.

Short-cycle Automation

Automated task operation in which manual control is alternated with short periods of

automation based on other operation environment elements, either the automation or the

human can instigate the change in control.

Situational Awareness

“The perception of the elements in the environment within a span of time and space, the

comprehension of their meaning and the projection of their status in the near future.”

(Endsley 1995)

Spatially Localized Cues

Auditory cues that are perceived to emanate from a specific location within a three

dimensional sound stage.

Take Control Message

Communication from the automation system to the driver requesting that the driver

resume manual control of the vehicle.

Transfer of Control

The transition between states of manual and automated control of the vehicle.

Tell-tale

An indicator which displays the status of a situation or system.

Transport Information and Control Systems (TICS)

Systems which improve the safety and efficiency of land-based transportation through the

use of automation and technology.

Two-point Threshold

The minimum separation distance needed for two distinct objects to be felt separately,

rather than as a single sensation, based off of two-point discrimination, which is an ability

to distinguish adjacent objects that are touching the skin as being independent objects.

Veiling Luminance

Uniform luminance that causes a reduction in contrast as it washes over the retina. It is

caused when the eye is exposed to a light source that is substantially more intense than

the adaptation level.

Vibrotactile Interface

An interface which provides information to the driver using physical vibrations of the

seat, seat belt, foot pedals, or steering wheel against the driver’s body.

Chapter 10. Index

Abbreviations (Chapter 11), 11-1

Accuracy, 5-11, 8-6, 8-7

Adaptive Automation, 1-3, 1-8

Adaptive Function Allocation, 1-3

Auditory Icons, 1-10, 4-4, 6-1

Auditory Interfaces (Chapter 6), 6-1

Auditory Messages, 4-2, 4-4 thru 4-7, 6-1, 6-3

Loudness, 4-2, 6-3, 6-6 (topic)

Perceived Annoyance of, 6-4 (topic)

Perceived Urgency of, 6-2 (topic)

Automation

Current Mode and Status, 3-2 (topic)

Driver Mental Models, 3-10 (topic)

Driver Training, 3-10, 3-11, 4-3, 49, 7-2 thru 4, 8-7

Driver Training Material, 3-14 (topic)

Etiquette, 3-16 (topic)

Messages, Display Properties, 3-4 (topic)

Special Considerations for Level 2 and Level 3, 3-12 (topic)

Transferring Control to Driver, 3-8 (topic)

Transferring Control to System, 3-6 (topic)

Cautionary Warnings, 5-9, 6-5 thru 7

Character Height for Icons and Text, 4-2, 5-10 (topic)

Compensatory Principle, 1-2

Control Placement, 8-3, 8-4 (topic)

Display Glare, 5-2, 5-4 (topic)

Distraction/Distracted Driving, 1-2,1-5, 3-1, 3-7, 4-1, 4-5, 4-6, 4-9, 5-6,

5-7, 5-13, 6-3, 6-5, 8-1, 8-2

Driver Comprehension, 4-5, 4-6, 4-8, 5-12, 6-1,

Designing Messages for (topic), 4-2

Driver Inputs (Chapter 8), 8-1

Driver-DVI Interactions, General Guidance, 8-2 (topic)

Dynamic Function Allocation, 1-3

Equations (Chapter 12), 12-1

Eye Glance, 8-2

False and Nuisance Alarms, 1-7, 5-13, 6-4, 6-5

Flashing Signals, 3-4, 3-12, 4-2, 5-1, 5-3, 5-12, 5-13

General Guidance for Level 2 and Level 3 Automation (Chapter 3), 3-1

Glossary (Chapter 9), 9-1

Haptic Displays/Signals, 3-4, 4-2, 4-4 thru 9, 5-9

Distinctiveness, 7-4 (topic)

Detectable Signal in a Vibrotactile Seat, 7-8 (topic)

Selection of, 7-2 (topic)

Vibrotactile Sensitivity Across the Body, 7-6 (topic)

Haptic Interfaces (Chapter 7), 7-1

Head-Up Displays (HUDs), 4-6 thru 8, 5-2, 5-3, 5-6 (topic), 5-13

High Head-Down Displays (HHDDs), 4-6, 4-8, 5-2, 5-3, 5-7

Imminent Collision Warning (ICW), 6-6, 6-7

Index (Chapter 10), 10-1

Lane Change Warning (LCW), 7-4

Lane Departure Warning (LDW), 1-4, 4-7, 7-4

Message Characteristics (Chapter 4), 4-1

Message Complexity, 4-4 (topic)

Mode Awareness Failure, 1-8, 3-7

Multimodal Messages, 3-16, 4-8 (topic)

Nomadic Devices/Systems, 8-1, 8-5

Redundancy, 3-16, 4-6, 4-8, 4-9, 7-5, 7-7

References (Chapter 14), 14-1

Relevant Documents From USDOT, SAE International, and ISO (Chapter 13), 13-1

Residual Function Principle, 1-2

Seat Pressure Distribution, 7-8, 7-9

Sensory Modality, Selection of, 4-6 (topic)

Situation Awareness, 1-12, 3-2, 3-12, 3-13

Speech Messages, 4-5, 4-8, 6-1, 6-3

Status Information, 1-2, 1-5, 1-6, 1-7, 1-8, 1-11, 1-12, 3-2 (topic), 3-6 thru 10, 3-12,

3-13, 3-16, 4-6, 5-7, 5-8, 8-3

Display Properties for, 3-4 (topic)

Types of Status Information, 3-2

Training Material for Automated Driving System Applications, 3-14 (topic)

Transitions, 1-1, 1-5, 1-16, 1-10, 3-2 thru 9, 3-12, 3-13, 4-5, 5-1, 5-7, 5-12, 8-1 thru

3, 8-6, 8-7

Urgency, 3-5, 3-8, 4-2, 5-8, 5-9, 5-12, 5-13, 6-1, 6-2 (topic), 64 thru 6-7

Visual Displays

Glare, 5-2, 5-4 (topic)

Locations, 4-6, 4-8, 5-1, 5-2 (topic), 5-4, 5-9

Using Color, 4-2, 5-1, 5-3, 5-8 (topic)

Temporal Characteristics, 4-2, 4-8, 4-9, 5-12 (topic), 7-1 thru 3, 7-5

Visual Interfaces (Chapter 5), 5-1

Voice Recognition Inputs, 8-5, 8-6 (topic)

Chapter 11. Abbreviations

AAM .................................................................................... Alliance of Automobile Manufacturers

ACAS ............................................................................... Automotive Collision Avoidance System

ACC ............................................................................................................ adaptive cruise control

ADAS .................................................................................... Advanced Driver Assistance System

AHS .................................................................................................... Automated Highway System

ANSI .................................................................................... American National Standards Institute

BSMS ................................................................................................ blind spot monitoring system

BSW ................................................................................................................... blind spot warning

CCW ................................................................................................... cautionary collision warning

cm .................................................................................................................................. centimeter

CSW ............................................................................................................. Curve Speed Warning

DARPA ....................................................... Defense Advanced Research Projects Administration

dB ......................................................................................................................................... decibel

dBA ................................................................................................................ decibel (A-weighted)

DVI ............................................................................................................ Driver-Vehicle-Interface

EID ...................................................................................................... Ecological Interface Design

FAA ............................................................................................... Federal Aviation Administration

FCW ........................................................................................... Forward Collision/Crash Warning

FHWA .......................................................................................... Federal Highway Administration

FMCSA ...................................................................... Federal Motor Carrier Safety Administration

FMVSS ............................................................................ Federal Motor Vehicle Safety Standards

FTA ................................................................................................. Federal Transit Administration

g ............................................................... acceleration/deceleration level (1 g = approx. 9.8 m/s

)

HDD .................................................................................................................. head-down display

HHDD ........................................................................................................ high head-down display

HMI ........................................................................................................ human-machine interface

HUD ....................................................................................................................... head-up display

HVAC ...................................................................................... heating, ventilating, air conditioning

Hz ........................................................................................................................................... Hertz

ICW ....................................................................................................... imminent collision warning

IP .......................................................................................................................... instrument panel

ISO .......................................................................... International Organization for Standardization

ITS ............................................................................... Intelligent Transportation Systems/Society

IVW .................................................................................................................... in-vehicle warning

L0 ..................................................................................................................... Level 0 Automation

L1 ..................................................................................................................... Level 1 Automation

L2 ..................................................................................................................... Level 2 Automation

L3 ..................................................................................................................... Level 3 Automation

L4 ..................................................................................................................... Level 4 Automation

L5 ..................................................................................................................... Level 5 Automation

LCDAS ................................................................................ Lane Change Decision Assist System

LCS ........................................................................................................... Lane Centering System

LCW ............................................................................................................... lane change warning

LDW ........................................................................................................... lane departure warning

LED ................................................................................................................... light-emitting diode

LHDD .......................................................................................................... low head-down display

ms ................................................................................................................................. millisecond

MT ....................................................................................................................... masked threshold

nm .................................................................................................................................. nanometer

NHTSA .................................................................. National Highway Traffic Safety Administration

OEM ............................................................................................ original equipment manufacturer

OOTL ....................................................................................................................... out-of-the-loop

PTT .............................................................................................................................. push-to-talk

RITA ............................................................. Research and Innovative Technology Administration

s .......................................................................................................................................... second

SAE ............................................................................................. Society of Automotive Engineers

SPL ................................................................................................................ sound pressure level

TICS ........................................................................... Transport Information and Control Systems

TTC ........................................................................................................................ time-to-collision

V2V ..................................................................................................................... vehicle-to-vehicle

VPT ............................................................................................................ variable priority training

Chapter 12. Equations

Glare Angle and Display Luminance





= 



 





+ 





  1 + 



62.5







Where:

veil

= Veiling luminance

glare

= Luminous intensity of glare source

θ = Glare angle

A = Driver age

5-4

Viewing Distance and Symbol Height

= 60  Arctan 



1000  



Viewing Distance and Visual Angle

= 1000   Tan 





Visual Angle and Symbol Height

=









Where:

H = Symbol height in millimeters

D = Viewing distance in meters (0.5-1.1m)

V = Visual angle subtended in arcminutes

5-10

100

Minimum Density of Vibrating Motors

2 x T

Where:

= Area width

= Area height

T = Two-point threshold

7-6

101

Chapter 13. Relevant Documents from the United States Department of

Transportation, SAE International, and International Organization

for Standardization

The table below provides an alphanumeric list of standards, best practices, and general resource

documents which may or may not be directly cited in chapter topics but, if not, present additional

information relevant to the design issues discussed in this document.

Document Description

ANSI/HFES 100-2007

Human factors engineering of computer

workstations

Year: 2007

Document Type: U.S. National Standard

Specifications for acceptable applications of human factors

engineering principles and practices related to computer workstation

design and configuration, defined as an operator-machine system

comprised of associated user-interface components (input devices,

output devices, and furniture).

ANSI Z535.3

American national standard. Criteria for

safety symbols

Year: 2011

Document Type: U.S. National Standard

General criteria for the design, evaluation, and use of safety symbols

for identifying and warning against specific hazards.

CIE 146

CIE equations for disability glare

Year: 2002

Document Type: International Standard

Definitions for disability glare equations for veiling luminance.

DOT 37-13

Visual-manual NHTSA driver distraction

guidelines for in-vehicle electronic devices

Year: 2013

Document Type: Federal Guidelines

Provides nonbinding, voluntary guidelines to discourage excessively

distracting devices in vehicles. Applicable to original equipment in-

vehicle electronic devices used for secondary tasks such as

communications, entertainment, information gathering, or navigation

through visual-manual means.

ISO 2575

Road vehicles--Symbols for controls,

indicators and tell-tales

Year: 2010

Document Type: International Standard

Specifications for symbols for use on controls, indicators and tell-

tales. Applicable to passenger cars, light and heavy commercial

vehicles, and buses. Also included are colors of possible optical tell-

tales for informing drivers of either correct operation or

malfunctioning of related devices.

ISO 3864-3

Graphical symbols--Safety colours and

safety signs--Part 3: Design principles for

graphical symbols for use in safety signs

Year: 2012

Document Type: International Standard

Principles, criteria and guidance for designing graphical symbols for

use in safety signs (described in ISO 3864‑1), and for the safety sign

element of product safety labels (described in ISO 3864‑2).

102

Document Description

ISO 4040

Road vehicles--Location of hand controls,

indicators and tell-tales in motor vehicles

Year: 2009

Document Type: International Standard

Specifications for the location of controls in motor vehicles, and for

certain combinations of functions for multifunction controls.

Applicable to hand-operated controls, indicators and tell-tales in all

motor vehicles, excluding motorcycles and mopeds, as defined in ISO

3833.

ISO 7731

Ergonomics--Danger signals for public and

work areas--Auditory danger signals

Year: 2003

Document Type: International Standard

Specification of criteria for the recognition of auditory danger signals,

especially in situations of high ambient noise.

ISO 9921

Ergonomics--Assessment of speech

communication

Year: 2003

Document Type: International Standard

Specification of requirements for the performance of speech

communication for verbal alert and danger signals, information

messages, and speech communication. Methods for

predicting/assessing subjective and objective performance in practical

applications are described. Examples are provided.

ISO 11429

Ergonomics--System of auditory and

visual danger and information signals

Year: 1996

Document Type: International Standard

Specification of warning and information signals which differentiate

between degrees of urgency, from extreme urgency to All Clear

situations.

ISO 15005

Road vehicles--Ergonomic aspects of

transport information and control systems--

Dialogue management principles and

compliance procedures

Year 2002

Document Type: International Standard

Description of ergonomic principles for designing dialogues between

drivers and the vehicle's transport information and control systems

(TICS) while the vehicle is in motion, including specification of

compliance verification conditions related to the principles.

ISO 15006

Road vehicles--Ergonomic aspects of

transport information and control systems--

Specifications for in-vehicle auditory

presentation

Year: 2011

Document Type: International Standard

Ergonomic specifications for auditory information displays related to

transport information and control systems (TICS), primarily when the

vehicle is in motion although it may also be applied when the vehicle

is stationary. Requirements and recommendations for in-vehicle

auditory signals from TICS, as well as characteristics and functional

factors for maximizing auditory signal intelligibility and utility while

helping prevent auditory or mental overload are provided.

ISO 15007-1

Road vehicles--Measurement of driver

visual behaviour with respect to transport

information and control systems--Part 1:

Definitions and parameters

Year: 2014

Document Type: International Standard

Definitions of key terms and parameters used in the analysis of driver

visual behavior for both real-world trials and laboratory-based driving

simulator studies.

103

Document Description

ISO 15008

Road vehicles--Ergonomic aspects of

transport information and control systems--

Specifications and test procedures for in-

vehicle visual presentation

Year: 2009

Document Type: International Standard

Specification of minimum requirements for image quality and

legibility of visual displays containing changeable information

presented while the vehicle is in motion, including test methods and

measurements for assessing compliance where necessary. Applicable

to mainly perceptual components of the visual information such as

character legibility and colour recognition; not applicable to factors

affecting performance and comfort such as coding, format and

dialogue characteristics, pictorial information or images, maps and

topographic representations.

ISO 15623

Transport information and control systems--

Forward vehicle collision warning systems--

Performance requirements and test

procedures

Year: 2002

Document Type: International Standard

Specification of performance requirements and test procedures for

systems which warn the driver of short inter-vehicle distance and

closing speed which could cause a rear-end collision with other

vehicles. Applicable to operations on roads with curve radii over

125m or higher radius curves.

ISO 16673

Road vehicles--Ergonomic aspects of

transport information and control systems--

Occlusion method to assess visual demand

due to the use of in-vehicle systems

Year: 2007

Document Type: International Standard

Procedure for measuring visual demand during the use of visual or

visual-manual interfaces while the vehicle is in motion. Applicable to

both original equipment and aftermarket in-vehicle systems.

ISO 17287

Road vehicles--Ergonomic aspects of

transport information and control systems--

Procedure for assessing suitability for use

while driving

Year: 2003

Document Type: International Standard

Procedure for assessing whether transport information and control

systems (TICS), or a combination of TICS with other in-vehicle

systems, are suitable for drivers while the vehicle is in motion. Topics

include user-oriented TICS description and context of use, TICS task

description and analysis, the assessment process, and documentation.

(Does not recommend specific variables for assessing suitability nor

defines criteria for establishing the suitability of use of a TICS table

while driving.)

ISO 17361

Intelligent transport systems--Lane

departure warning systems--Performance

requirements and test procedures

Year: 2007

Document Type: International Standard

Specifications for the definition of the system, classification,

functions, human-machine interface (HMI) and test methods for in-

vehicle lane departure warning systems which are appropriate for

highways and highway-like roads; warnings at roadway sections

having temporary or irregular lane markings (such as roadwork zones)

is not within the scope. Applicable for systems which provide

warnings only (no automatic mitigation action) for passenger cars,

commercial vehicles and buses.

104

Document Description

ISO 17386

Transport information and control systems-

Maneuvering aids for low speed operation

(MALSO) Performance requirements and

test procedures

Year: 2010

Document Type: International Standard

Specification of minimum functionality requirements for light-duty

vehicles, e.g., passenger cars, pickup trucks, light vans and sport

utility vehicles (motorcycles excluded) equipped with MALSO

systems such as detection of and information on the presence of

relevant obstacles within a defined (short) detection range. It defines

minimum requirements for failure indication as well as performance

test procedures; it includes rules for the general information strategy

but does not restrict the kind of information or display system.

(Sensing technology is not addressed; visibility-enhancement systems

without distance ranging and warning are not addressed. For reversing

aids and obstacle-detection devices on heavy commercial vehicles,

see ISO/TR 12155.)

ISO 17387

Intelligent transport systems--Lane change

decision aid systems (LCDAS)--

Performance requirements and test

procedures

Year: 2008

Document Type: International Standard

Specification of system requirements and test methods for Lane

Change Decision Aid Systems (LCDAS) and their use on forward

moving cars, vans and straight trucks in highway situations (does not

address LCDAS for use on motorcycles or articulated vehicles).

ISO 26022

Road vehicles--Ergonomic aspects of

transport information and control systems--

Simulated lane change test to assess in-

vehicle secondary task demand

Year: 2010

Document Type: International Standard

An estimate of secondary task demand derived through a laboratory

base-method, which quantitatively measures human performance

degradation on a primary driving-like task while a secondary task is

being performed. The method, for both original equipment and

aftermarket in-vehicle systems, applies to in-vehicle information,

communication, entertainment, control, manual, visual, haptic and

auditory single and combination systems for passenger cars, but

cannot be used to test secondary tasks requiring that speed variations

be performed.

ISO 26262

Road vehicles--Functional safety

Year: 2011

Document Type: International Standard

Intended to be applied to electrical and/or electronic systems in

production passenger vehicles (excludes heavy vehicles). The

standard covers functional safety across the product lifecycle,

provides an automotive-specific risk-based approach for determining

risk arising from malfunctioning electrical/electronic systems, and

provides requirements for validating and confirming measures to

ensure sufficient and acceptable level of safety.

ISO/TR 12204

Road vehicles--Ergonomic aspects of

transport information and control systems--

Introduction to integrating safety critical

and time critical warning signals

Year: 2012

Document Type: Technical Report

General, informational guidance for integration of safety critical and

time critical warning signals into existing in-vehicle messages

presented to a driver; does not provide guidance in integration of non-

critical signals, nor how to design an integrated warning HMI.

ISO/TR 16352

Road vehicles--Ergonomic aspects of in-

vehicle presentation for transport

information and control systems--Warning

systems

Year: 2005

Document Type: Technical Report

Literature survey of human-machine interface of warning systems in

vehicles which discusses efficiency, acceptance of different

modalities and combinations of warnings, and design parameters of

visual, auditory and tactile warnings.

105

Document Description

ISO/TS 14198

Road vehicles--Ergonomic aspects of

transport information and control systems--

Calibration tasks for methods which assess

driver demand due to the use of in-vehicle

systems

Year: 2012

Document Type: Technical Specification

Procedures for developing secondary, calibration tasks used in a dual-

task setting for assessing drivers’ attentional demand when using in-

vehicle systems. Advice provided for selecting an appropriate

candidate calibration task and includes its application, experimental

design, data collection, and procedures for analysis.

ISO/TS 15007-2

Road vehicles--Measurement of driver

visual behaviour with respect to transport

information and control systems--Part 2:

Equipment and procedures

Year: 2001

Document Type: Technical Specification

Guidelines for analyzing driver visual behaviour to assist in planning

evaluation trials, specifying/installing data capture equipment, as well

as analyzing, interpreting and reporting visual-behaviour

measurement. Applicable to road trials and simulated driving

environments, but is not applicable to the assessment of head-up

displays.

ISO/TS 16951

Road vehicles--Ergonomic aspects of

transport information and control systems

(TICS)--Procedures for determining

priority of on-board messages presented to

drivers

Year: 2004

Document Type: Technical Specification

Provides formal procedures and two alternate methods for

determining the priority of in-vehicle messages, including traveler,

navigation, traffic advisories, warnings, system status and other

information, as well as messages from other sources such as

telephones, warnings, and tell-tales.

MIL-STD-1472G

Department of defense design criteria

standard - Human engineering

Year: 2012

Document Type: Military Standard

General human engineering design criteria for military systems,

subsystems, equipment, and facilities with the intent of optimal

system performance given inherent human capabilities and

limitations.

SAE J287

Driver hand control reach

Year: 2007

Document Type: Recommended Practice

Description of the boundaries of hand control locations that can be

reached by a percentage of different driver populations in passenger

cars, multi-purpose passenger vehicles, and light trucks (Class A

vehicles); not applicable to heavy trucks (Class B vehicles).

SAE J941

Motor vehicles drivers’ eye locations

Year: 2010

Document Type: Recommended Practice

Establishment of the location of drivers’ eyes inside a vehicle for

passenger cars, multi-purpose passenger vehicles, and light trucks

(Class A vehicles) and heavy trucks (Class B vehicles) (eyellipses

have not been updated from previous versions of SAE J941).

SAE J1138

Design criteria--Driver hand controls

location for passenger cars, multipurpose

passenger vehicles, and trucks (10,000

GVW and under)

Year: 2009

Document Type: Recommended Practice

Description of design criteria related to the location and labeling of

hand controls (does not include hand-held devices such as remote

controls or cellular phones).

106

Document Description

SAE J1757-1

Standard metrology for vehicular displays

Year: 2007

Document Type: Standard

Methods to determine optical performance for Flat Panel Displays in

all typical automotive ambient light illumination, focusing on High

Ambient Contrast Ratio, a critical element for display legibility in a

sunshine environment.

SAE J2364

Navigation and route guidance function

accessibility while driving

Year: 2004

Document Type: Recommended Practice

Establishment of both a static method and an interrupted vision

method for determining which navigation and route guidance

functions should be accessible to the driver while the vehicle is in

motion; applicable to original equipment and aftermarket route-

guidance system functions for passenger vehicles. Does not apply to

visual monitoring tasks which do not require a manual control input,

such as route following, nor to voice-activated controls or passenger

operation of controls.

SAE J2365

Calculation of the time to complete in-

vehicle navigation and route guidance

tasks

Year: 2002

Document Type: Recommended Practice

A method for calculating the time needed to complete navigation

system-related tasks which may be used as an assessment tool for

safety and usability of alternative navigation and route guidance

system interfaces. Does not consider voice-activated controls, voice

output from the navigation system, communication between the driver

and others, or passenger operation. Applicable to both original

equipment and aftermarket route-guidance and navigation system

functions for passenger vehicles.

SAE J2395

ITS in-vehicle message priority

Year: 2002

Document Type: Recommended Practice

Description of a method for prioritizing ITS in-vehicle messages

and/or displayed information which is applicable to original

equipment and aftermarket ITS message-generating systems for

passenger vehicles and heavy trucks. A prioritization value is assigned

to specific messages or units of information which is then used to

determine the order that simultaneous, or overlapping, in-vehicle

messages are presented to the driver.

SAE J2396

Definitions and experimental measures

related to the specification of driver visual

behavior using video based techniques

Year: 2000

Document Type: Recommended Practice

Description of key terms and metrics applied in the analysis of video-

based driver eye glance behavior, intended to assist development of a

common source of reference for driver visual behavior data. Data

collated and analyzed from this document allow comparisons to be

performed across different device evaluations and experimental

scenarios.

SAE J2399

Adaptive cruise control (ACC) operating

characteristics and user interface

Year: 2003

Document Type: Standard

Specifications for the minimum requirements for Adaptive Cruise

Control system operating characteristics and elements of the user

interface; applicable to original equipment and aftermarket ACC

systems for passenger vehicles (including motorcycles). Not

applicable to commercial vehicles nor variations on ACC, such as

“stop-and-go” ACC.

SAE J2400

Human factors in forward collision

warning systems: Operating characteristics

and user interface requirements

Year: 2003

Document Type: Information Report

Description of the elements for a Forward Collision Warning user

interface, and requirements and test methods for these systems.

Applicable to original equipment and aftermarket FCW systems for

passenger vehicles including cars, light trucks, and vans, but does not

apply to heavy trucks, nor does it address integration issues associated

with adaptive cruise control (ACC).

107

Document Description

SAE J2402

Road vehicles—Symbols for controls,

indicators, and tell-tales

Year: 2010

Document Type: Standard

Specification of symbols for use on controls, indicators, and tell-tales

which is applicable to passenger cars, light and heavy commercial

vehicles, and buses.

SAE J2678

Navigation and route guidance function

accessibility while driving rationale

Year: 2004

Document Type: Recommended Practice

Description of the rationale used by the Navigation Function

Accessibility Subcommittee for the development and content of the

SAE J2364 Recommended Practice: Navigation and Route Guidance

Function Accessibility While Driving.

SAE J2802

Blind spot monitoring system (BSMS):

Operating characteristics and user interface

Year: 2010

Document Type: Recommended Practice

Specification of the minimum recommendations for Blind Spot

Monitoring System (BSMS) operational characteristics and elements

of the user interface. Applicable to original equipment and

aftermarket BSMS systems for passenger vehicles, but not

motorcycles or heavy trucks, nor does it address Lane Change

Systems (which monitor a larger area behind the vehicle). A visual

BSMS indicator is recommended.

SAE J2808

Road/lane departure warning systems:

Information for the human interface

Year: 2007

Document Type: Information Report

Recommendations for Road Departure Warning Systems (RDWS)

operational characteristics and elements of the user interface.

Applicable to original equipment and aftermarket systems for light-

duty vehicles on relatively straight roads with a radius of curvature of

500m or more, and under good weather conditions.

SAE J2830

Process for comprehension testing of in-

vehicle icons

Year: 2008

Document Type: Information Report

A process for testing driver comprehension of safety, navigation,

infotainment or other ITS message symbols or icons.

SAE J2831

Development of design and engineering

recommendations for in-vehicle

alphanumeric messages

Year: 2012

Document Type: Information Report

Recommendations for alphanumeric messages communicated to the

vehicle by external (e.g., RDS, satellite radio) or internal

(e.g., infotainment system) sources while the vehicle is in-motion.

Applicable to OEM (embedded) and aftermarket systems. Does not

cover ergonomic issues regarding display characteristics such as

viewing angle, brightness, contrast, font design, etc.

SAE J3016

Taxonomy and definitions for terms related

to on-road motor vehicle automated

driving systems

Year: 2014

Document Type: Information Report

Taxonomy and operational definitions for the full range of levels of

automation in on-road motor vehicles as a foundation for discussion

and further standards development within the

“Automated/Autonomous Vehicle” community.

108

Chapter 14. References

49 C.F.R. pt. 571.208. Federal Motor Vehicle Safety Standard (FMVSS) No. 208, Occupant

crash protection [Docket No. 74-14; Notice 103], RIN 2127-AG14, Retrieved from

www.nhtsa.dot.gov/cars/rules/rulings/Labels5.mlv.html

Abe, G., & Richardson, J. (2004). The effect of alarm timing on driver behaviour: an

investigation of differences in driver trust and response to alarms according to alarm

timing. Transportation Research Part F: Traffic Psychology and Behaviour, 7(4-5), 307-

22.

Abe, G., & Richardson, J. (2006). The influence of alarm timing on driver response to collision

warning systems following system failure. Behaviour & Information Technology, 25(5),

443-452.

Adrian, W., & Bhanji, A. (1991). Fundamentals of disability glare: A formula to describe stray

light in the eye as a function of glare angle and age. Proceedings of the First International

Symposium on Glare (pp. 185-193). New York: Lighting Research Institute.

Aldridge, L. C., & Lansdown, T. C. (1999). Driver preferences for speech based interaction with

in-vehicle systems. Proceedings of the Human Factors and Ergonomics Society Annual

Meeting, 977-981.

Alliance of Automobile Manufacturers (AAM). (2006). Statement of principles, criteria and

verification procedures on driver interactions with advanced in-vehicle information and

communication systems, including 2006 updated sections [Report of the Driver Focus-

Telematics Working Group]. Retrieved from

www.autoalliance.org/index.cfm?objectid=D6819130-B985-11E1-9E4C000C296BA163.

Arroyo, E., Sullivan, S., & Selker, T. (2006). CarCoach: A polite and effective driving coach.

Proceedings of CHI ’06 Extended Abstracts on Human Factors in Computing Systems,

357-362.

Baldwin, C. L., & May, J. F. (2011). Loudness interacts with semantics in auditory warnings to

impact rear-end collisions. Transportation Research Part F: Traffic Psychology and

Behaviour, 14(1), 36-42.

Bhise, V. D. (2011). Ergonomics in the automotive design process. Boca Raton, FL: CRC Press.

Blanco, M., Atwood, J., Vasquez, H. M., Trimble, T. E., Fitchett, V. L., Radlbeck, J., ... Morgan,

J. F. (2015, August). Human factors evaluation of level 2 and level 3 automated driving

concepts (Report No. DOT HS 812 182). Washington, DC: National Highway Traffic Safety

Administration. Available at

www.nhtsa.gov/sites/nhtsa.dot.gov/files/812182_humanfactorseval-l2l3-

automdrivingconcepts.pdf

Boot, W. R., Basak, C., Erickson, K. I., Neider, M., Simons, D. J., Fabiani, M., . . . Kramer, A. F.

(2010). Transfer of skill engendered by complex task training under conditions of variable

priority. Acta Psychologia, 135, 349-357.

Brock, J. F., McFann, J., Inderbitzen, R. E., & Bergoffen, G. (2007). Effectiveness of commercial

motor vehicle driver training curricula and delivery methods (CTBSSP Synthesis MC-13).

Washington, DC: National Academy Press.

109

Brookhuis, K. A., van Driel, C.J.G., Hof, T., van Arem, B., & Hoedemaeker, M. (2008). Driving

with a congestion assistant; mental workload and acceptance. Applied Ergonomics, 40,

1019-1025. doi: 10.1016/j.apergo.2008.06.010

Brown, S. B., Lee, S. E., Perez, M. A., Doerzaph, Z. R., Neale, V. L., & Dingus, T. A. (2005).

Effects of haptic brake pulse warnings on driver behavior during an intersection approach.

Proceedings of the Human Factors and Ergonomics Society 49th Annual Meeting, 1892-

1896.

Campbell, J. L., Bittner, A. C., Jr., Lloyd, M., Mitchell, E., & Everson, J. H. (1997). Driver-

vehicle interface (DVI) design guidelines for the intersection collision avoidance (ICA)

system. Seattle, WA: Battelle Human Factors Transportation Center.

Campbell, J. L., Carney, C., & Kantowitz, B. H. (1998). Human factors design guidelines for

advanced traveler information systems (ATIS) and commercial vehicle operations (CVO)

(Report No. FHWA-RD-98-057). Washington, DC: Federal Highway Administration.

Campbell, J. L., Richard, C.M., Brown, J. L., & McCallum, M. (2007). Crash warning system

interfaces: Human factors insights and lessons learned, final report (Report No. DOT HS

810 697). Washington, DC: National Highway Traffic Safety Administration. Retrieved

from

www.nhtsa.gov/DOT/NHTSA/NRD/Multimedia/PDFs/Crash%20Avoidance/2007/CWS_

HF_Insights_Task_5_Final_Rpt.pdf

Campbell, J. L., Richman, J. B., Carney, C., & Lee, J. D. (2004). In-vehicle display icons and

other information elements, Volume I: Guidelines (Report No. FHWA-RD-03-065).

McLean, VA: Federal Highway Administration. Retrieved from

www.fhwa.dot.gov/publications/research/safety/03065/index.cfm

Carney, C., Campbell, J. L., & Mitchell, E. A. (1998). In-vehicle display icons and other

information elements: Literature review (Report No. FHWA-RD-98-164). Washington,

DC: Federal Highway Administration.

Carsten, O., Lai, F. C. H., Barnard, Y., Jamson, A. H., & Merat, N. (2012). Control task

substitution in semiautomated driving: Does it matter what aspects are automated? Human

Factors: The Journal of the Human Factors and Ergonomics Society, 54(5), 747-761. doi:

10.1177/0018720812460246.

Cassidy, A. M. (2009). Mental models, trust, and reliance: Exploring the effect of human

perceptions on automation use (Master’s thesis). Naval Postgraduate School, Monterey,

CA.

Catchpole, K. R., McKeown, J. D., & Withington, D. J. (2004). Localizable auditory warning

pulses. Ergonomics, 47(7), 748-771.

Chan, A. H. S., & Ng, A. W. Y. (2009). Perceptions of implied hazard for visual and auditory

alerting signals. Safety Science. 47, 346-352.

Charissis, V., & Papanastasiou, S. (2010). Human-machine collaboration through vehicle head

up display interface. Cognition, Technology and Work, 12(1), 41-50.

Cholewiak, R. W., & Craig, J. C. (1984). Vibrotactile pattern recognition and discrimination at

several body sites. Perception and Psychophysics, 35(6), 503-514.

CIE . (2002). CIE 146:2002:CIE equations for disability glare. Vienna, Austria: International

Commission on Illumination.

110

Cullinane, B., & Kirn, C. (2012). Timely comprehension: A methodology for evaluating

integration of in-vehicle warning systems. Presented at SAE 2012 World Congress &

Exhibition. doi:10.4271/2012-01-0100.

Cummings, M. L., Mastracchio, C., Thornburg, K. M., & Mkrtchyan, A. (2013). Boredom and

distraction in multiple unmanned vehicle supervisory control. Interacting with Computers,

25(1), 34-47.

De Groot, S., de Winter, J. C. F., Garcia, J. M., Mulder, M., & Wieringa, P. A. (2011). The effect

of concurrent bandwidth feedback on learning the lane-keeping task in a driving simulator.

Human Factors, 53(1), 50-62.

Deatherage, B. H. (1972). Auditory and other sensory forms of information presentation. In H. P.

Van Cott & R. G. Kinkade (Eds.), Human engineering guide to equipment design (rev. ed.)

(pp. 123-160). Washington, DC: U.S. Government Printing Office.

Dobbins, T., & McKinley, A. (2008). Integration issues of tactile displays in military

environments. In J. B. F. van Erp & B. P. Self (Eds.), Tactile displays for orientation,

navigation and communication in air, sea and land environments (Chapter 5) (Report No.

RTO-TR-HFM-122). NATO Science and Technology Organization. Retrieved from

www.cso.nato.int/pubs/rdp.asp?RDP=RTO-TR-HFM-122.

Dorneich, M. C., Ververs, P. M., Mathan, S., Whitlow, S., & Hayes, C. C. Considering etiquette

in the design of an adaptive system. Journal of Cognitive Engineering and Decision

Making, 6 (2), 243-265.

Doshi, A., Cheng, S. U., & Trivedi, M. M. (2009). A novel active heads-up display for driver

assistance. IEEE Transactions on Systems, Man, and Cybernetics—Part B: Cybernetics,

39(1), 85-93.

Easterby, R. S. (1970). The perception of symbols for machine displays. Ergonomics, 13(1), 149-

158.

Edworthy, J., & Hellier, E. (2000). Auditory warnings in noisy environments. Noise Health, 2(6),

27-40.

Edworthy, J., Loxley, S., & Dennis, I. (1991). Improving auditory warning design: Relationship

between warning sound parameters and perceived urgency. Human Factors, 33(2), 205-

231.

Endsley, M. R. (1995). Towards a theory of situation awareness in dynamic systems. Human

Factors, 37, 32-64.

Endsley, M. R. (2012). Situation awareness. In Salvendy, G. (Ed.), Handbook of Human Factors

and Ergonomics, Fourth Edition (Chapter 19). Hoboken, NJ: John Wiley & Sons.

Endsley, M. R., & Kaber, D. B. (1999). Level of automation effects on performance, situation

awareness and workload in a dynamic control task. Ergonomics, 42(3), 462-492.

Endsley, M. R., & Kiris, E. O. (1995). The out-of-the-loop performance problem and level of

control in automation. Human Factors, 37(2), 381-394.

Farber, E., & Matle, C. (1989). PCDETECT: A revised version of the DETECT seeing distance

model. Transportation Research Record: Journal of the Transportation Research Board,

1213, 11-20.

111

Federal Aviation Administration. (2013). Manual flight operations (Safety Alert for Operators

[SAFO] 13002). Retrieved from the FAA website at

www.faa.gov/other_visit/aviation_industry/airline_operators/airline_safety/safo/all_safos/

media/2013/SAFO13002.pdf

Federal Highway Administration. (2012). Manual on uniform traffic control devices, 2009

edition with revisions 1 and 2. Washington, DC: Author.

Fitch, G. M., Hankey, J. M., Kleiner, B. M., & Dingus, T. A. (2011). Driver comprehension of

haptic seat alerts intended for use in an integrated collision avoidance system.

Transportation Research Part F: Traffic Psychology and Behaviour, 14, 278-290.

Fitch, G. M., Kiefer, R. J., Hankey, J. M., & Kleiner, B. M. (2007). Toward developing an

approach for alerting drivers to the direction of a crash threat. Human Factors, 49(4), 710-

720.

Fitts, P. M., (Ed.). (1951). Human engineering for an effective air navigation and traffic control

system. Washington, DC: National Research Council.

Flemisch, F., & Schieben, A. (Eds.) (2009). Validation of preliminary design by simulation.

Retrieved from the Highly Automated Vehicles for Intelligent Transport (HAVEit) website

at http://haveit-

eu.org/LH2Uploads/ItemsContent/24/HAVEit_212154_D33.3_Superfinal.pdf

Forkenbrock, G., Snyder, A., Heitz, M., Hoover, R. L., O'Harra, B., Vasko, S., & Smith, L.

(2011). A test track protocol for assessing forward collision warning driver-vehicle

interface effectiveness (Report No. DOT HS 811 501). Washington, DC: National Highway

Traffic Safety Administration.

Fricke, N. (2007). Effects of adaptive information presentation. Proceedings of the 4th

International Driving Symposium on Human Factors in Driver Assessment, Training and

Vehicle Design, 292-298.

Gallace, A., Ho, T.Z., & Spence, C. (2007). Multisensory numerosity judgments for visual and

tactile stimuli. Perception & Psychophysics, 69, 487-501.

Gallace, A., Tan, H. Z., & Spence, C. (2007). The body surface as a communication system: The

state of the art after 50 years. Presence, 16 (6), 655-676.

Gellatly, A. W., & Dingus, T. A. (1998). Speech recognition and automotive applications: Using

speech to perform in-vehicle tasks. Proceedings of the Human Factors and Ergonomics

Society Annual Meeting, 1247-1251.

General Motors Corporation & Delphi-Delco Electronic Systems. (2002). Automotive collision

avoidance system field operation test, warning cue implementation summary report (Report

No. DOT HS 809 462). Washington, DC: National Highway Traffic Safety Administration.

General Motors Corporation. (2005). Automotive collision avoidance system field operational

test (ACAS FOT) final program report (Report No. DOT HS 809 886). Washington, DC:

National Highway Traffic Safety Administration.

Giddings, B.J. (1972). Alpha-numerics for raster displays. Ergonomics, 15(1), 65-72.

Gold, C., Damböck, D., Lorenz, L., & Bengler, K. (2013). “Take over!” How long does it take to

get the driver back into the loop? Proceedings of the Human Factors and Ergonomics

Society 57th Annual Meeting, 1938-1942.

112

Gold, C., Lorenz, L., Damböck, D., Bengler, K. (2013, November). Partially automated driving

as a fallback level of high automation. Paper presented at Tagung Fahrerassistenzsysteme:

Der weg zum automatischen fahren. Retrieved from

http://mediatum.ub.tum.de/doc/1187198/1187198.pdf

Gonzalez, C., Lewis, B. A., Roberts, D. M., Pratt, S. M., & Baldwin, C. L. (2012). Perceived

urgency and annoyance of auditory alerts in a driving context. Proceedings of the Human

Factors and Ergonomics Society Annual Meeting, 56(1), 1684-1687.

Goodrich, M. A.., & Boer, E. R. (1999). Multiple mental models, automation strategies, and

intelligent vehicle systems. Proceedings of the IEEE/IEEJ/JSAI Conference on Intelligent

Transportation Systems. Tokyo, Japan.

Goodrich, M. A., & Boer, E. R. (2003). Model-based human-centered task automation: A case

study in ACC system design. IEEE Transactions on Systems, Man, and Cybernetics, Part

A, Systems and Humans, 33(3), 325-336.

Goodrich, M. A., Boer, E. R., & Inoue, H. (1999). A model of human brake initiation behavior

with implications for ACC design. Proceedings of IEE/IEEJ/JASI International Conference

on Intelligent Transportation Systems, 86-91.

Graving, J. S., Easterlund, P. A., & Manser, M. M. (2011). Developing a bus driver training

program for a driver assistive system. Proceedings of the Human Factors and Ergonomics

Society 55

Annual Meeting, 1543-1547.

Grote, G., Weik, S., Wäfler, T., & Zölch, M. (1995). Complementary allocation of functions in

automated work systems. In Y. Anzai, K. Ogawa, & H. Mori (Eds.), Symbiosis of Human

and Artifact. Amsterdam: Elsevier.

Guilluame, A., Drake, C., Rivenez, M., Pellieux, L., & Chastres, V. (2002). Perception of

urgency and alarm design. Proceedings of the 8th International Conference on Auditory

Display.

Hancock, P. A. (2007). On the process of automation transition in multitask human-machine

systems. IEEE Transactions on Systems, Man, and Cybernetics, Part A: Systems and

Humans, 37(4), 586-598.

Hecht, D., & Reiner, M. (2009). Sensory dominance in combinations of audio, visual and haptic

stimuli. Experimental Brain Research, 193, 307-314.

Hellier, E., & Edworthy, J. (1989). Quantifying the perceived urgency of auditory warnings.

Canadian Acoustics, 17(4), 3-11.

Higuchi, M., & Raksincharoensak, P. (2010). Seat vibrotactile warning interface for forward

vehicle collision avoidance. Proceedings of the SICE Annual Conference 2010, 1370-1373.

Hilburn, B., Molloy, R., Wong, D., & Parasuraman, R. (1993). Operator versus computer control

of adaptive automation. Proceedings of the 7th International Symposium on Aviation

Psychology (pp. 161-166). Columbus, OH: The Ohio State University.

Ho, C., Tan, H. Z., & Spence, C. (2005). Using spatial vibrotactile cues to direct visual attention

in driving scenes. Transportation Research Part F: Traffic Psychology and Behaviour,

8(6), 397-412.

Hoc, J.-M., Young, M. S., & Blosseville, J.-M. (2009). Cooperation between drivers and

automation: Implications for safety. Theoretical Issues in Ergonomics Science, 10(2), 135-

160.

113

Hoeger, R., Zeng, H., Hoess, F., Kranz, T., Boverie, S., Straus, M., Jakobsson, E., . . . Nilsson,

A. (2011). The future of driving. Deliverable D61.1. Final report. Version 1.0 (Report for

the Highly Automated Vehicles for Intelligent Transport [HAVEit] 7th Framework

Programme). Retrieved from www.haveit-

eu.org/displayITM1.asp?ITMID=24&LANG=EN.

Hoffman, J. D., Lee, J. D., McGehee, D. V., Macias, M., & Gellatly, A. W. (2005). Visual

sampling of in-vehicle text messages: Effects of number of lines, page presentation, and

message control. Transportation Research Record: Journal of the Transportation Research

Board, 1937, 22-30.

Hogema, J. H., Sjoerd, C.D.V., Van Erp, J., & Kiefer, R. J. (2009). A tactile seat for direction

coding in car driving: Field evaluation. IEEE Transaction on Haptics, 2(4), 181-188.

Hollnagel, E. (2004). Automation and human work. In C. Sandom & R. S. Harvey, (Eds.),

Human Factors for Engineers. London: Institute of Engineering and Technology.

Hollnagel, E., & Bye, A. (2000). Principles for modeling function allocation. International

Journal of Human-Computer Studies, 52, 253-265.

Horrey, W. J., Wickens, D. D., & Alexander, A. L. (2003). The effects of head-up display clutter

and in-vehicle display separation on concurrent driving performance. Proceedings of the

Human Factors and Ergonomics Society 47th Annual Meeting. 1880-1884.

Horvitz, E. (1999). Principles of mixed-initiative user interfaces. Proceedings of CHI ’99, ACM

SIG-CHI Conference on Human Factors in Computing Systems, Pittsburgh, PA.

Howell, W. C., & Kraft, C. L. (1959). Size, blur, and contrast as variables affecting the legibility

of alphanumeric symbols on radar-type displays (WADC Technical Report 59-536).

Wright-Patterson Air Force Base, OH: Wright Air Development Center (DTIC No. AD-

232 889).

Human Factors and Ergonomics Society. (2007). ANSI/HFES 100-2007: Human factors

engineering of computer workstations. Santa Monica, CA: Author.

Inagaki, T. (2003). Adaptive automation: Sharing and trading of control. In E. Hollnagel (Ed.),

Handbook of Cognitive Task Design. Mahwah, NJ: Erlbaum.

International Organization for Standardization. (1984). Development and principles for

application of public information symbols. (ISO/TR 7239). Geneva: Author.

International Organization for Standardization. (2005). Road vehicles—Ergonomic aspects of in-

vehicle presentation for transport information and control systems—Warning systems

(ISO/TR 16352). Geneva: Author.

ISO. (2002). ISO 15005: Road vehicles—Ergonomic aspects of transport information and

control systems—Dialogue management principles and compliance procedures. Geneva:

International Organization for Standards.

ISO. (2009). ISO 15008: Road vehicles—Ergonomic aspects of transport information and

control systems—Specifications and test procedures for in-vehicle visual presentation.

Geneva: International Organization for Standardization.

ISO. (2008). ISO 17387: Intelligent transport systems—Lane change decision aid systems

(LCDAS)—Performance requirements and test procedures. Geneva: International

Organization for Standardization.

114

ISO. (1988). ISO 3461-1: General principles for the creation of graphical symbols, Part I:

Graphical symbols for use on equipment. Geneva: International Organization for

Standardization.

ISO. (2003). ISO 7731: Danger signals for public and work areas—auditory danger signals.

Geneva: International Organization for Standardization.

Ivancic, K., & Hesketh, B. (2000). Learning from errors in a driving simulation: Effects on

driving skill and self-confidence. Ergonomics, 43(12), 1966-1984.

Jamson, A.H., Merat, N., Carsten, O.M.J., & Lai, F.C.H. (2013). Behavioural changes in drivers

experiencing highly-automated vehicle control in varying traffic conditions. Transportation

Research Part C: Emerging Technologies, 30, 116-125.

Ji, Y. J., Lee, L., & Hwang, W. (2011). Haptic perception in the vehicle seat. Human Factors

and Ergonomics in Manufacturing and Service Industries, 21(3), 305-325.

Jones, C. M., Gray, R., Spense, C., & Tan, H. Z. (2008). Directing visual attention with spatially

informative and spatially noninformative tactile cues. Experimental Brain Research,

186(4), 659-669.

Jones, L. A., & Sarter, N. B. (2008). Tactile displays: Guidance for their design and application.

Human Factors, 50(1), 90-111.

Kaber, D. B., & Endsley, M. R. (1997). Out-of-the-loop performance problems and the use of

intermediate levels of automation for improved control system functioning and safety.

Process Safety Progress, 16(3), 126-131.

Kaber, D. B., & Endsley, M. R. (2004). The effects of level of automation and adaptive

automation on human performance, situation awareness and workload in a dynamic control

task. Theoretical Issues in Ergonomics Science, 5(2), 113-153. doi:

10.1080/1463922021000054335.

Kaber, D. B., Riley, J. M., Tan, K.-W., & Endsley, M. R. (2001). On the design of adaptive

automation for complex systems. International Journal of Cognitive Ergonomics, 5(1), 37-

57.

Kandel, E. R., Schwartz, J. H., & Jessell, T. M. (2000). Principles of neural science (4th Ed.).

New York: McGraw-Hill Health Professions.

Kazi, T. A., Stanton, N. A., Walker, G. H., & Young, M. S. (2007). Designer driving: Drivers’

conceptual model and level of trust in adaptive cruise control. International Journal of

Vehicle Design, 45(3), 339-360.

Kiefer, R. J. (1996). A review of driver performance with head-up displays. Third World

Congress on Intelligent Transport Systems [CD-ROM]. Washington, DC: ITS America.

Kiefer, R. J., Cassar, M. T., Flannagan, C. A., Jerome, C. J., & Palmer, M. D. (2005). Surprise

braking trials, time-to-collision judgments, and “first look” maneuvers under realistic

rear-end crash scenarios (Forward Collision Warning Requirements Project, Tasks 2 and

3a Final Report (Report No. DOT HS 809 902). Washington, DC: National Highway

Traffic Safety Administration.

Kiefer, R., LeBlanc, D., Palmer, M., Salinger, J., Deering, R., & Shulman, M. (1999).

Development and validation of functional definitions and evaluation procedures for

collision warning/avoidance systems (Report No. DOT HT 808 964). Washington, DC:

National Highway Traffic Safety Administration.

115

Kun, A., Paek, T., Medenica, Z. (2007). The effect of speech interface accuracy on driving

performance. Proceedings of the 8th Annual Conference of the International Speech

Communication Association (Interspeech 2007), 2332-2335.

Laurienti, P. J., Burdette, J. H., Maldjian, J. A., & Wallace, M. T. (2006). Enhanced multisensory

integration in older adults. Neurobiology of Aging, 27, 1155-1163.

Lee, J. D., McGehee, D. V., Brown T. L., & Marshall, D. (2006). Effects of adaptive cruise

control and alert modality on driver performance. Transportation Research Record:

Journal of the Transportation Research Board, 1980, 49-56.

Lee, J. D., McGehee, D. V., Brown, T. L., & Marshall, D. (2006). Effects of adaptive cruise

control and alert modality on driver performance. Transportation Research Record:

Journal of the Transportation Research Board, 1980, 49-56.

Lee, J. D., McGehee, D. V., Brown, T. L., & Marshall, D. (2006). Effects of adaptive cruise

control and alert modality on driver performance. Transportation Research Record:

Journal of the Transportation Research Board, 1980, 49-56.

Lee, J. D., McGehee, D. V., Brown, T. L., & Nakamoto, J. (2012). Driver sensitivity to brake

pulse duration and magnitude. Ergonomics, 50(6), 828-836.

Lee, J. D., McGehee, D. V., Brown, T. L., & Reyes, M. L. (2002). Driver distraction, warning

algorithm parameters, and driver response to imminent rear-end collisions in a high-

fidelity driving simulator (Report No. DOT HS 809 448). Washington DC: National

Highway Traffic Safety Administration.

Lee, J. D., & Moray, N. (1994). Trust, self-confidence, and operators’ adaptation to automation.

International Journal of Human-Computer Studies, 40, 153-184.

Lee, J. D., & See, K. A. (2004). Trust in automation: Designing for appropriate reliance. Human

Factors, 46(1), 50-80.

Lee, J. D.., & Seppelt, B. D. (2012). Human factors and ergonomics in automation design. In G.

Salvendy (Ed.), Handbook of Human Factors and Ergonomics, Fourth Edition (pp.1615-

1642). Hoboken: Wiley & Sons.

Lee, S. E., Knipling, R. R., DeHart, M. C., Perez, M. A., Holbrook, G. T., Brown, S. B., . . .

Olson, R. L. (2004). Vehicle-based countermeasures for signal and stop sign violations:

Task 1. Intersection control violation crash analyses. (Report No. DOT HS 809 716).

Washington, DC: National Highway Traffic Safety Administration.

Leonard, S. D. (1999). Does color of warning affect risk perception? International Journal of

Industrial Ergonomics, 23, 499-504.

Lerner, N. D., Kotwal, B. M., Lyons, R. D., & Gardner-Bonneau, D. J. (1996). Preliminary

human factors guidelines for crash avoidance warning devices. Interim report (Report No.

DOT HS 808 342). Washington, DC: National Highway Traffic Safety Administration.

Available at https://archive.org/stream/preliminaryhuman00lern#page/n1/mode/2up

Levitan, L., Bunus, M., Dewing, W. L., Reinhart, W., Vora, P. B., & Llaneras, R. E. (1997).

Preliminary human factors guidelines for automated highway system designers (2nd Ed.)

(Report No. FHWA-RD-97-125). McLean, VA: Federal Highway Administration.

Lind, H. (2007). An efficient visual forward collision warning display for vehicles. SAE

Technical Paper 2007-01-1105. doi:10.4271/2007-01-1105.

116

Llaneras, R. E., Salinger, J., & Green, C. A. (2013). Human Factors issues associated with

limited ability autonomous driving systems: Drivers' allocation of visual attention to the

forward roadway. Proceedings of the International Driving Symposium on Human Factors

in Driver Assessment, Training, and Vehicle Design, 7, 92-98.

Lo, V. E-W., & Green, P. A. (2013). Development and evaluation of automotive speech

interfaces: Useful information from the human factors and related literature. International

Journal of Vehicular Technology, 2013, Article ID 924170. Retrieved from

www.hindawi.com/journals/ijvt/2013/924170/.

Lumsden, J. (2008). Handbook of research on user interface design and evaluation for mobile

technology, Volume I. Retrieved from www.comm.rwth-aachen.de/files/lumsden122.pdf

Marshall, D. C., Lee, J. D., & Austria, P. A. (2007). Alerts for in-vehicle information systems:

Annoyance, urgency, and appropriateness. Human Factors, 49(1), 145-157.

Martens, M., Pauwelussen, J., Schieben, A., Flemisch, F., Merat, N., Jamson, A. H., & Caci, R.

(2008). Human factors’ aspects in automated and semi-automatic transport systems: State

of the art (Report 3.2.1). Brussels: European Commission.

McDougall, S., Tyrer, V., & Folkard, S. (2006). Searching for signs, symbols, and icons: Effects

of time of day, visual complexity, and grouping. Journal of Experimental Psychology,

12(2), 118-128. doi: 10.1037/1076-898X.12.2.118.

McGuirl, J. M., & Sarter, N. B. (2006). Supporting trust calibration and the effective use of

decision aids by presenting dynamic system confidence information. Human Factors,

48(4), 656–665.

McKnight, J., & Adams, B. (1970). Driver education and task analysis, Volume 1: Task

descriptions (Report No. DOT HS 800 367). Washington DC: National Highway Traffic

Safety Administration.

Merat, N., & Jamson, A. H. (2009). How do drivers behave in a highly automated car?

Proceedings of the International Driving Symposium on Human Factors in Driver

Assessment, Training, and Vehicle Design, 5, 514-521.

Michon, J. A. (1985). A critical view of driver behavior models: What do we know, what should

we do? In L. Evans & R. C. Schwing (Eds.) Human Behavior and Traffic Safety (pp. 485-

520). New York: Plenum Press.

Miller, C. A., & Funk, H. B. (2001). Associated with etiquette: Meta-communication to make

human-automation interaction more natural, productive, and polite. Proceedings of the 8th

European Conference on Cognitive Science Approaches to Process Control. Munich,

Germany.

Miller, C. A., & Parasuraman, R. (2007). Designing for flexible interaction between humans and

automation: Delegation interfaces for supervisory control. Human Factors, 49(1), 57-75.

Department of Defense. (January 11, 2012). MIL-STD-1472G. Department of Defense Design

Criteria Standard: Human Engineering. Washington, DC: U.S. Government Printing

Office.MIL-STD-1472G (2012). Human engineering. Washington, DC: Department of

Defense. Available at http://everyspec.com/MIL-STD/MIL-STD-1400-

1499/download.php?spec=MIL-STD-1472G.039997.pdf.

Mourant, R. R., & Langolf, G. D. (1976). Luminance specifications for automobile instrument

panels. Human Factors, 18(1), 71-84.

117

Muir, B. M., & Moray, N. (1996). Trust in automation: Part II. Experimental studies of trust and

human intervention in a process control simulation. Ergonomics, 39, 429-460.

NASA Ames Research Center Color Usage Research Lab. (n.d.) Using color in information

display graphics. Retrieved from the NASA Ames Research Center Color Usage Research

Lab website at http://colorusage.arc.nasa.gov/index.php.

National Highway Traffic Safety Administration. (2011). Crash warning interface metrics,

Phase 2 (Report No. DOT HS 811 471). Washington, DC: Author.

National Highway Traffic Safety Administration. (2013). Preliminary statement of policy

concerning automated vehicles (NHTSA 14-13). Retrieved from

www.nhtsa.gov/About+NHTSA/Press+Releases/U.S.+Department+of+Transportation+Rel

eases+Policy+on+Automated+Vehicle+Development.

Norman, D. A. (1990). The problem of automation: Inappropriate feedback and interaction, not

over-automation. Philosophical Transactions of the Royal Society of London, B, 1-18.

O’Day, S., & Tijerina, L. (2011). Legibility: Back to the basics. SAE International Journal of

Passenger Cars—Mechanical Systems, 4(1), 591-604.

O’Toole, R. (2009). Gridlock: Why we’re stuck in traffic and what to do about it. Washington,

DC: Cato Institute Press.

O'Hanlon, J. F., & Kelley, G. R. (1977). Comparison of performance and physiological changes

between drivers who perform well and poorly during prolonged vehicular operation. In R.

R. Mackie (Ed.), Vigilance: Theory, Operational Performance and Physiological

Correlates (pp. 87-109). New York: Plenum Press.

Owens, J. M., McLaughlin, S. B., & Sudweeks, J. (2011). Driver performance while text

messaging using handheld and in-vehicle systems. Accident Analysis & Prevention 43,

939-947. doi: doi:10.1016/j.aap.2010.11.019.

Panou, M. C., Bekiaris, E. D., & Touliou, A. A. (2010). ADAS module in driving simulation for

training young drivers. Proceedings of the 13th International IEEE Annual Conference on

Intelligent Transportation Systems, 1582-1587.

Parasuraman, R., Hilburn, B., Molloy, R., & Singh, I. (1991). Adaptive automation and human

performance: III. Effects of practice on the benefits and costs of automation shifts

(NAWCADWAR-92037-60). Warminster, PA: Naval Air War Center-Aircraft Division.

Parasuraman, R., & Miller, C. A. (2004). Trust and etiquette in high-criticality automated

systems. Communications of the ACM, 47(4), 51-55.

Parasuraman, R., & Riley, V. (1997). Humans and automation: Us, misuse, disuse, abuse.

Human Factors, 39(2), 230-253.

Parasuraman, R., Sheridan, T. B., & Wickens, C. D. (2000). A model of types and level of

human interaction with automation. IEEE Transactions on Systems, Man and Cybernetics,

Part A, Systems, and Humans, 30(3), 286-297.

Parasuraman, R., & Wickens, C. D. (2008). Humans: Still vital after all these years of

automation. Human Factors, 50(3), 511-520.

Perez, M. A., Kiefer, R. J., Haskins, A., & Hankey, J. M. (2009). Evaluation of forward collision

warning system visual alert candidates and SAE J2400. SAE International Journal of

Passenger Cars- Mechanical Systems, 2(1), 750-764. doi:10.4271/2009-01-0547

118

Pickrell, T. M., & Liu, C. (2014). Seat belt use in 2013: Overall results (Report No. DOT HS

811 875). Washington, DC: National Highway Traffic Safety Administration.

Pollock, E., Chandler, P., & Sweller, J. (2002). Assimilating complex information. Learning and

Instruction, 12, 61-68.

Pomerleau, D., Jochem, T., Thorpe, C., Batavia, P., Pape, D., Hadden, J., . . . Everson, J. (1999).

Run-off-road collision avoidance using IVHS countermeasures, Final report (Report No.

DOT HS 809 170). Washington, DC: National Highway Traffic Safety Administration.

Popp, M. M., & Faerber, B. (1993). Feedback modality for nontransparent driver control actions:

Why not visually? In A.G. Gale, I.D., Brown, C.M. Haslegrave, H.W. Kryusse, & S.P.

Taylor (Eds.), Vision in Vehicles—IV (pp.263-270). Amsterdam: Elsevier Science.

Porter, M. M. (2013). Older driver training using video and global positioning system

technology—a randomized controlled trial. Journals of Gerontology Series A: Biomedical

Sciences and Medical Sciences, 68(5), 574-580. doi:10.1093/Gerona/gls160.

Prinzel, L. J., III, & Risser, M. (2004). Head-up displays and attention capture (NASA/TM-

2004-213000). Retrieved from the NASA Technical Reports Server website at

http://naca.larc.nasa.gov/search.jsp?R=20040065771&qs=N%3D4294966788%2B4294724

598%2B4294323833.

Rajaonah, B., Anceaux, F., & Vienne, F. (2006). Trust and use of adaptive cruise control: A

study of a cut-in situation. Cognition, Technology and Work, 8(2), 146-155.

Rasmussen, J. (1983). Skills, rules, and knowledge: Signals, signs, and symbols, and other

distinctions in human performance models. IEEE Transactions on Systems, Man, and

Cybernetics, SMC-13(3), 257-266. doi: 10.1109/TSMC.1983.6313160.

Redding, R. E., Cannon, J.R., & Seamster, T. L. (1992). Expertise in air traffic control (ATC):

What is it, and how can we train for it? Proceedings of the Human Factors and Ergonomics

Society, 36, 1326-1330.

Reiner, A. (2010). Sensor-actuator supported implicit interaction in driver assistance systems.

Wiesbaden, Germany: Vieweg+Teubner.

Rosario, H., Louredo, M., Diaz, I., Soler, A., Gil, J. J., Solaz, J. S., & Jordi, J. (2010). Efficacy

and feeling of a vibrotactile frontal collision warning implemented in a haptic pedal.

Transportation Research Part F: Traffic Psychology and Behaviour, 13(2), 80-91.

Rudin-Brown, C. M., & Parker, H. A. (2004). Behavioural adaptation to adaptive cruise control

(ACC): Implications for preventative strategies. Transportation Research Part F: Traffic

Psychology and Behaviour, 7, 59-76.

Ryu, J., Chun, J., Choi, S., & Han, S. H. (April, 2010). Vibrotactile feedback for information

delivery in the vehicle. IEEE Transactions on Haptics, 3(2), 138-149.

Society of Automotive Engineers. (2009). SAE J1138: Design criteria—Driver hand controls

location for passenger cars, multipurpose passenger vehicles, and trucks (10,000 GVW and

under). Warrendale, PA: SAE International.

SAE. (2003). SAE J2399: Adaptive cruise control (ACC) operating characteristics and user

interface. Warrendale, PA: SAE International.

SAE. (2003). SAE J2400: Human factors in forward collision warning systems: Operating

characteristics and user interface requirements. Warrendale, PA: SAE International.

119

SAE. (2008). SAE J2830: Process for comprehension testing of in-vehicle icons. Warrendale,

PA: SAE International.

SAE. (2007). SAE J267: Driver hand control reach. Warrendale, PA: SAE International.

SAE. (2014). SAE J3016: Taxonomy and definitions for terms related to on-road motor vehicle

automated driving systems. Warrendale, PA: SAE International.

Salinger, J. (2012). Human factors for limited-ability autonomous driving systems. Retrieved from

http://onlinepubs.trb.org/onlinepubs/conferences/2012/Automation/presentations/

Salinger.pdf

Sarter, N. B., & Woods, D. D. (1995). How in the world did we ever get into that mode? Mode

error and awareness in supervisory control. Human Factors, 37, 5-19. doi:

10.1518/001872095779049516.

Sarter, N. B., Woods, D. D., & Billings, C. E. (1997). Automation surprises. In G. Salvendy

(Ed.), Handbook of Human Factors and Ergonomics, Second Edition (pp. 1926-1943).

Hoboken: Wiley & Sons.

Scott, J. J., & Gray, R. (2008). A comparison of tactile, visual, and auditory warnings for rear-

end collision prevention in simulated driving. Human Factors, 50(2), 264-275. doi:

10.1518/001872008X250674

Seppelt, B. D., & Lee, J. D. (2007). Making adaptive cruise control (ACC) limits visible.

International Journal of Human-Computer Studies, 65, 192-205.

Sheridan, T. B. and Parasuraman, R. (2005). Human-automation interaction. Reviews of Human

Factors and Ergonomics, 1, 89-129.

Soccolich, S. A., Blanco, M., Hanowski, R. J., Olson, R. L., Morgan, J. F., Guo, F., & Wu, S. C.

(2013). An analysis of driving and working hour on commercial motor vehicle driver safety

using naturalistic data collection. Accident Analysis & Prevention, 58, 249-258.

Spain, R. D., & Madhavan, P. (2009). The role of automation etiquette and pedigree in trust and

dependence. Proceedings of the Human Factors and Ergonomics Society 53

Annual

Meeting, 339-343.

Spence, C., Ngo, M. K., Lee, J.-H., & Tan, H. (2010). Solving the correspondence problem in

haptic/multisensory interface design. In M. H. Zadeh (Ed.). Advances in Haptics (pp. 47-

74). Available from: www.intechopen.com/books/advances-in-haptics/solving-

thecorrespondence-problem-in-haptic-multisensory-interface-design.

Stanley, L. M. (2006). Haptic and auditory cues for lane departure warnings. Proceedings of the

Human Factors and Ergonomics Society Annual Meeting, 50, 2405-2408. doi:

10.1177/154193120605002212.

Stanton, N. A., Dunoyer, A., & Leatherland, A. (2011). Detection of new in-path targets by

drivers using stop & go adaptive cruise control. Applied Ergonomic, 42, 592-601.

Stanton, N. A., & Marsden, P. (1996). From fly-by-wire to drive-by-wire: Safety implications of

vehicle automation. Safety Science, 24, 35-49.

Stanton, N. A., & Young, M. S. (1998). Vehicle automation and driving performance.

Ergonomics, 41(7), 1014-1028.

Stanton, N. A., & Young, M. S. (2000). A proposed psychological model of driving automation.

Theoretical Issues in Ergonomics Science, 1(4), 315-331.

120

Stanton, N. A., Young, M. S., Walker, G. H., Turner, H., & Randle, S. (2001). Automating the

driver’s control tasks. International Journal of Cognitive Ergonomics, 5(3), 221-236.

Stevens, A., Hallen, H., Pauzie, A., Vezier, B., Gelau, C., Eckstein, L., . . . Höfs, W. (2005).

European statement of principles on the design of human machine interaction (ESoP 2005,

Draft). Directorate-General Information Society Technologies and Media of the European

Commission.

Strickland, D. (2013, August 13). Every co-pilot needs a pilot. Washington Automotive Press

Association Luncheon. Washington, DC.

Suied, C., Susini, P., & McAdams, S. (2008). Evaluating warning sound urgency with reaction

times. Journal of Experimental Psychology: Applied, 14(3), 201-212.

Suzuki, K., & Jansson, H. (2003) An analysis of driver’s steering behavior during auditory or

haptic warnings for designing of lane departure warning system. JSAE Review, 24, 63-70.

Tan, A. K., & Lerner, N. D. (1995). Multiple attribute evaluation of auditory warning signals for

in-vehicle crash avoidance warning systems (Report No. DOT HS 808 535). Washington,

DC: National Highway Traffic Safety Administration, Office of Crash Avoidance

Research.

Thiffault, P., & Bergeron, J. (2003). Fatigue and individual differences in monotonous simulated

driving. Personality and Individual Differences, 34(1), 159-179.

Tijus, C., Barcenilla, J., Cambon de Lavalette, B., & Maunier, J.-G. (2007). The design,

understanding and usage of pictograms. In G. Rijlaarsdam (Series Ed.) Vol. 21. Studies in

Writing: Written Documents in the Workplace (pp. 17-31). Netherlands: Kluwer.

Toffetti, A. Wilschut, E. S., Martens, M. H., Schieben, A., Rambaldini, A., Merat, N., &

Flemisch, F. (2009). CityMobil: Human factor issues regarding highly automated vehicles

on eLane. Transportation Research Record: Journal of the Transportation Research

Board, 2110, 1-8. doi: 10.3141/2110-01.

Tsao, H.-S. J., Hall, R. W., & Shadlover, S. E. (1993). Design options for operating fully

automated highway systems. Berkeley: University of California PATH Institute of

Transportation Studies.

United States Army Training and Doctrine Command (TRADOC). (2003). Objective force

embedded training (OFET): User’s functional description (TRADOC Pamphlet No. 350-

37). Fort Monroe, VA: Author.

Wäfler, T., Grote, G., Windischer, A., & Ryser, C. (2003). KOMPASS: A method for

complementary system design. In E. Hollnagel (Ed.), Handbook of Cognitive Task Design.

Mahwah, NJ: Erlbaum.

Wickens, C. D., Hutchins, S., Carolan, T., & Cumming, J. (2013). Effectiveness of part-task

training and increasing-difficulty training strategies: A meta-analysis approach. Human

Factors, 55, 461-470.

Widmann, G., Salinger, J., Dufour, R., and Green, C. (2011). Human factors studies for limited-

ability autonomous driving systems (LAADS). Presentation at the SAE 2011

Government/Industry Meeting. Washington DC. Retrieved from

www.sae.org/events/gim/presentations/2011.

Wiener, E. (1989). Human factors of advanced technology (“glass cockpit”) transport aircraft

(NASA Contractor Report 177528). Moffett Field, CA: NASA Ames Research Center.

121

Wiener, E. L., Chute, R. D., & Moses, J. H. (1999). Transition to glass: Pilot training for high-

technology transport aircraft (NASA/CR-1999-208784). National Aeronautics and Space

Administration: Washington, DC.

Wiese, E. E., & Lee, J. D. (2004). Auditory alerts for in-vehicle information systems: The effects

of temporal conflict and sound parameters on driver attitudes and performance.

Ergonomics, 47(9), 965-986. doi: 10.1080/00140130410001686294.

Wilde, G. J. S. (1998). Risk homeostasis theory: An overview. Injury Prevention, 4, 89-91.

doi: 10.1136/ip.4.2.89.

Wilkison, B.D. (2008). Effects of mental model quality on collaborative system performance

(Master’s Thesis). Georgia Institute of Technology, Atlanta, GA.

Williges, B. H., & Williges, R. C. (1982). Structured human/computer dialogue using speech

technology. Proceedings of the Workshop on Standardization for Speech I/O Technology

(pp. 143-151). Gaithersburg, MD: National Bureau of Standards.

Wreggit, S. S., Powell, J., Kirn, C., & Hayes, E. (2000). Human factors evaluation of the truck

productivity computer’s electroluminescent display (Paper No. 2000-01-3409).

Warrendale, PA: SAE International.

Yoo, H., Tsimhoni, O., Wantanabe, H., Green, P., & Shah, R. (1999). Display of HUD warnings

to drivers: Determining an optimal location (UMTRI-99-9). Ann Arbor: University of

Michigan Transportation Research Institute.

Young, M. S., & Stanton, N. A. (1997). Automotive automation: Investigating the impact on

drivers’ mental workload. International Journal of Cognitive Ergonomics, 1, 325-336.

Young, M. S., & Stanton, N. A. (2007). Back to the future: Brake reaction times for manual and

automated vehicles. Ergonomics, 50(1), 46-58. doi: 10.1080/00140130600980789.

DOT HS 812 555

August 2018

13494-080318-v3