Handbook of Human Factors for Automated, Connected, and Intelligent Vehicles

Donald L. Fisher, William J. Horrey, John D. Lee, Michael A. Regan, Donald L. Fisher, William J. Horrey, John D. Lee, Michael A. Regan

• Publisher: CRC Press
• Year: 2020
• Print ISBN: 9781138035027
• eBook ISBN: 9781351979801
• Topic: Computer Science
• Subtopic: Human-Computer Interaction
• Index: Computer Science

Handbook of Human Factors for Automated, Connected, and Intelligent Vehicles

Subject Guide: Ergonomics & Human Factors

Automobile crashes are the seventh leading cause of death worldwide, resulting in over 1.25 million deaths yearly. Automated, connected, and intelligent vehicles have the potential to reduce crashes significantly, while also reducing congestion, carbon emissions, and increasing accessibility. However, the transition could take decades. This new handbook serves a diverse community of stakeholders, including human factors researchers, transportation engineers, regulatory agencies, automobile manufacturers, fleet operators, driving instructors, vulnerable road users, and special populations. It provides information about the human driver, other road users, and human–automation interaction in a single, integrated compendium in order to ensure that automated, connected, and intelligent vehicles reach their full potential.

Features

• Addresses four major transportation challenges—crashes, congestion, carbon emissions, and accessibility—from a human factors perspective
• Discusses the role of the human operator relevant to the design, regulation, and evaluation of automated, connected, and intelligent vehicles
• Offers a broad treatment of the critical issues and technological advances for the designing of transportation systems with the driver in mind
• Presents an understanding of the human factors issues that are central to the public acceptance of these automated, connected, and intelligent vehicles
• Leverages lessons from other domains in understanding human interactions with automation
• Sets the stage for future research by defining the space of unexplored questions

1 Introduction

Donald L. Fisher

Volpe National Transportation Systems Center

William J. Horrey

AAA Foundation for Traffic Safety

John D. Lee

University of Wisconsin Madison

Michael A. Regan

University of New South Wales

CONTENTS

Key Points

1.1 Background

1.2 Definitions

1.2.1 Levels of Automation and Active Safety Systems

1.2.1.1 Levels of Automation

1.2.1.2 Active Safety Systems

1.2.2 Automated, Connected, and Intelligent Vehicles

1.2.2.1 Automated Vehicles

1.2.2.2 Connected Vehicles

1.2.2.3 Intelligent Vehicles

1.2.3 Operational Design Domain

1.3 The Handbook: A Quick Guide

1.3.1 The State of the Art: ACIVs (Chapter 2)

1.3.2 Issues in the Deployment of ACIVs (Problems)

1.3.2.1 Driver’s Mental Model of Vehicle Automation (Chapter 3)

1.3.2.2 Driver Trust in ACIVs (Chapter 4)

1.3.2.3 Public Opinion about ACIVs (Chapter 5)

1.3.2.4 Workload, Distraction, and Automation (Chapter 6)

1.3.2.5 Situation Awareness in Driving (Chapter 7)

1.3.2.6 Allocation of Function to Humans and Automation and the Transfer of Control (Chapter 8)

1.3.2.7 Driver Fitness in the Resumption of Control (Chapter 9)

1.3.2.8 Driver Capabilities in the Resumption of Control (Chapter 10)

1.3.2.9 Driver State Monitoring for Decreased Fitness to Drive (Chapter 11)

1.3.2.10 Behavioral Adaptation (Chapter 12)

1.3.2.11 Distributed Situation Awareness (Chapter 13)

1.3.2.12 Human Factors Issues in the Regulation of Deployment (Chapter 14)

1.3.3 Human-Centered Design of ACIVs (Solutions)

1.3.3.1 HMI Design for ACIVs (Chapter 15)

1.3.3.2 HMI Design for Fitness Impaired Populations (Chapter 16)

1.3.3.3 Automated Vehicle Design for People with Disabilities (Chapter 17)

1.3.3.4 Importance of Training for ACIVs (Chapter 18)

1.3.4 Special Topics

1.3.4.1 Connected Vehicles in a Connected World: A Sociotechnical Systems Perspective (Chapter 19)

1.3.4.2 Congestion and Carbon Emissions (Chapter 20)

1.3.4.3 Automation Lessons from Other Domains (Chapter 21)

1.3.5 Evaluation of ACIVs

1.3.5.1 Human Factors Considerations in Testing and Evaluating ACIVs (Chapter 22)

1.3.5.2 Techniques for Making Sense of Behavior in Complex Datasets (Chapter 23)

1.4 Conclusion

Acknowledgments

References

Key Points

• Automated, connected, and intelligent vehicles hold great promise—increasing safety for all and mobility for underserved populations while decreasing congestion and carbon emissions.
• There may be unintended consequences of advances in automated technologies that affect the benefit that drivers can derive from these technologies, potentially slowing the development of the technologies themselves.
• Many of these unintended consequences center around human factors issues, issues between the driver and the vehicle, other road users, and the larger transportation system.
• Human factors research can be used to identify and seek to explain the unintended consequences, to develop and evaluate countermeasures, and to decrease, if not entirely avoid, any delay in the deployment of these technologies.

1.1 Background

We as humans cannot help but wonder what the future will hold and how it will unfold in time. When it comes to the effect of advanced technologies on our behaviors and on the behavior of the vehicles that we drive, the public speculation has been especially intense over the last ten years, starting in 2009 when Google¹ began its self-driving car program, now called Waymo. Such vehicles have the potential to substantially reduce the number of crashes, the level of carbon emissions, the congestion in our road systems, and the spread of wealth inequality, while at the same time increasing substantially opportunities for those who are mobility impaired (National Highway Traffic Safety Administration, 2017; Department of Transportation, 2019; Chang, 2015). Although some individuals are skeptical about early presumptions regarding the benefits of automated, connected, and intelligent vehicles (ACIVs) (Noy, Shinar, & Horrey, 2017; Bonnefon, Shariff, & Rahwan, 2016), the introduction of vehicles with advanced features continues to increase exponentially. As with the advent of the smartphone, anticipating the long-term positive and negative consequences of new technology is nearly impossible (e.g., Islam & Want, 2014; Twenge, Martin, & Campbell, 2018). It may be some time before we actually know the real benefits of such vehicles and features.

However, it is possible to take the bumps out of the road to full automation even without knowing the long-term consequences. This Handbook will focus specifically on the changes that will be wrought and the corresponding human factors challenges that need to be addressed by advances in the autonomy, connectivity, and intelligence of the vehicles that are being introduced into the fleet today and are likely to be introduced over the next several years. For readers relatively new to the discussion of why human factors concerns might be relevant to advanced technologies in the automobile, a simple example from one of the editors’ and authors’ long list of examples might help. This particular editor was driving 60 mph on a highway with two travel lanes in each direction and for a brief second or two fell asleep (had what is technically referred to as a “microsleep”). He drifted into the adjacent lane, woke up, and returned to his own lane. Had there been a large truck overtaking him in the adjacent lane, he might not be here to tell the story. Others’ lives may have been destroyed as well. But, fortunately, there was no truck and all was well. This speaks directly to the lifesaving potential of technologies which, in this case, could have kept the car in the lane and maintained speed adaptively. But it also points out just how beguiling these technologies can be.

Most vehicles on the road today that keep the car centered and adjust the speed require the driver to constantly monitor the driving environment (SAE International, 2018). Why? If we consider just automatic steering, there are many situations in which it may unexpectedly deactivate. The driver really does need to be in the loop. But, we also know that, perversely, automation can make it easier for the driver to fall out of the loop and become disengaged (Endsley, 2017; Endsley, 1995). Are we just trading off situations in which the technologies can be lifesaving for situations in which the technologies actually create conditions that increase the likelihood that a driver will crash if the technology cannot handle a particular scenario? This is the fundamental paradox of automation. While it can provide unparalleled opportunities, it comes with its own set of challenges.

Perhaps this paradox is best exemplified by a recent study of driver’s trust in automation. A field study was run in which the drivers were asked to navigate a network of roads on a closed course using a vehicle with both automatic steering and adaptive cruise control (ACC) (Victor, Tivestan, Gustafsson, Sangberg, & Aust, 2018). The drivers were told that they needed to monitor the driving environment and were warned by the vehicle driver state monitoring system if they did not comply. At the end of the drive, either a car or a large garbage bag was placed in their path. Both were easily crushed (e.g., a balloon car), but not obviously so to the driver before striking them. Driver’s trust in automation was measured after the drive on a scale of 0 (no trust) to 7 (high trust). Fully 21 of 76 drivers crashed (28%). All of the drivers who crashed had trust scores of 5 or higher (Victor, 2019). In short, the drivers became so reliant on the technology that they assumed it would avoid obstacles even when the technology encountered situations it was not designed to accommodate.

¹ Now Alphabet.

For readers familiar with the ongoing issues, you will find material in this Handbook which we believe will help set the stage for a human focus on future discussions about ACIVs. To date, the definition that vehicle manufacturers and major standards organizations have put forth concerning automation defines it primarily from a vehicle-centric point of view: as the technology capabilities increase, so too does the level of automation of the vehicle (SAE International, 2018). But this can easily mislead drivers into believing that their role in the driving process decreases as the levels of automation increase, despite warnings to the contrary, to the point where drivers actually feel that they can safely disengage from driving for long periods of time. A more driver-centric viewpoint is essential to extending automation safety benefits, one which defines and supports the new roles that drivers face.

The goal of this Handbook is to identify the real gains that can be achieved by identifying the various human factors, challenges, and opportunities that ACIVs pose and then, whenever possible, to suggest countermeasures (and opportunities for needed research), recognizing that the rapid advances in technology are changing both the challenges and the countermeasures. It is arguably the case that, even if these challenges were not addressed, there would be a net benefit to incorporating advanced technologies into new automobiles (Najm & daSilva, 2000), especially active safety systems like automatic emergency braking (AEB) and electronic stability control. But it is also arguably the case that, if these challenges are addressed, the net benefits will only increase. Moreover, by addressing these challenges, one reduces the real likelihood that the development of the technologies will be hobbled, if not halted, by crashes such as th...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Table of Contents
• Preface
• Editors
• Contributors
• Chapter 1 Introduction
• Chapter 2 Automated Driving: Decades of Research and Development Leading to Today’s Commercial Systems
• Chapter 3 Driver’s Mental Model of Vehicle Automation
• Chapter 4 Driver Trust in Automated, Connected, and Intelligent Vehicles
• Chapter 5 Public Opinion About Automated and Self-Driving Vehicles: An International Review
• Chapter 6 Workload, Distraction, and Automation
• Chapter 7 Situation Awareness in Driving
• Chapter 8 Allocation of Function to Humans and Automation and the Transfer of Control
• Chapter 9 Driver Fitness in the Resumption of Control
• Chapter 10 Driver Capabilities in the Resumption of Control
• Chapter 11 Driver State Monitoring for Decreased Fitness to Drive
• Chapter 12 Behavioral Adaptation
• Chapter 13 Distributed Situation Awareness and Vehicle Automation: Case Study Analysis and Design Implications
• Chapter 14 Human Factors Considerations in Preparing Policy and Regulation for Automated Vehicles
• Chapter 15 HMI Design for Automated, Connected, and Intelligent Vehicles
• Chapter 16 Human–Machine Interface Design for Fitness-Impaired Populations
• Chapter 17 Automated Vehicle Design for People with Disabilities
• Chapter 18 Importance of Training for Automated, Connected, and Intelligent Vehicle Systems
• Chapter 19 Connected Vehicles in a Connected World: A Sociotechnical Systems Perspective
• Chapter 20 Congestion and Carbon Emissions
• Chapter 21 Automation Lessons from Other Domains
• Chapter 22 HF Considerations When Testing and Evaluating ACIVs
• Chapter 23 Techniques for Making Sense of Behavior in Complex Datasets
• Chapter 24 Future Research Needs and Conclusions