Non-Motorized Behavior Changes and Communication with Autonomous Vehicles
With Autonomous Vehicles (AV) being introduced
into cities and intermixing with human-driven cars, non-motorized corridor
users (pedestrians and bicyclists) must interact with a computer driven car
instead of a human driver. Non-motorized users at times rely on eye contact,
hand motions, or audible dialogue with human drivers to accomplish roadway
crossings. The elimination of human interaction and communication with AV
technology could influence the behavior of non-motorized users, introducing
unpredictability with the elimination of the human element in the driver seat. Additional
concerns about the lack of human interaction and vulnerable road users such as
the visually impaired or children who rely on the interaction to validate the
safety of their behaviors or human drivers to react to unpredictable behavior
strengthen the need for enhanced communication between AV and non-motorized
users to create predictable behavior.
Identifying training strategies, tools and
applications that can guide and influence non-motorized user behavior could
reduce unpredictability and improve safety by minimizing the human factor in
the interaction, and enhance communication between AV and pedestrians or
bicyclists. Many non-motorized users have hand held devices (smart phones) that
could provide a means for communication or advanced warning. External features
on the AV or the behavior of AV may be used to influence behavior.
Perceiving and predicting pedestrian behavior is
a critical piece of the pedestrian-AV interaction, suggesting that models of
pedestrian behavior may feed into the protocol for contextual warnings or
messages for non-motorized users.
Understanding the differences in information
requirement based on the environment is also critical. Context, such as traffic
volume, number of lanes, setting, socio-economic status, gender, age, access to
technology, and number of non-motorized users could influence behavior and
communication. Features may be context-dependent: an audible warning, for
example, might be helpful on a rural road at a greenway crossing, but might be
useless in a dense urban environment where warnings would occur all the time.
Identifying context-sensitive opportunities in the built environment could
guide non-motorized user behavior in the interaction with AV as well as provide
public agencies with planning guidelines for future facilities that facilitate
positive non-motorized user behavior with AV interactions.
The objectives of this
research are as follows:
what is needed for non-motorized users to make a decision when interacting
with a vehicle.
current research on effectiveness of warning systems on hand held devices and
study influence of warning systems on non-motorized user behavior.
options for communicating with hand held devices not currently in use and
influences to the non-motorized user behavior.
the interaction of vulnerable users and human-driven vehicles to determine
communication and influences on behavior and identify behaviors that could
influence AV technology development.
effectiveness of warning systems on vulnerable users
influence of external features and vehicle behavior on non-motorized user
built environment features that positively influence non-motorized user
behavior in AV interactions.
training strategies to educate non-motorized user behavior to improve safety
when interacting with AV technology.
the differences in information requirements based on context, such as
urban/rural settings, traffic volume, number of lanes, and number of
predictive models of pedestrian behavior and how they might inform warning or
effects of communication and no communication to non-motorized user behavior.
Most of the current or previous research on AV technology has focused on
the built environment, the capabilities of AV technology communicating with
built environment aspects or other technology to alert other corridor users to
the presence of the AV. This study is expected to identify the human factors
that influence non-motorized user behavior. It is also expected to identify the
effectiveness of communication tools used by non-motorized users to communicate
with vehicles and the resulting influence on non-motorized user behavior. The
study of the built environment and influences on non-motorized user behavior in
AV interactions may benefit public agencies as those agencies identify
infrastructure updates. The study of vulnerable users communication and
behavior addresses potential safety concerns and provides equity.
Planning for Walking and Cycling in an Autonomous Vehicle
Future, Mid Atlantic Transportation Sustainability University Transportation
Center, DTRT13-G-UTC33, 2018
Performance Assessment & Simulation of Pedestrian Behavior at
Unsignalized Crossings, Research and Innovative Technology Administration, 98215,
Perceptions and Willingness to Interact with Autonomous Vehicles, US DOT, 69A355174125,
Human-Machine Communication in the Context of Autonomous Vehicles, US DOT,
69A3551747115, 2017, Ongoing
Audience: AV technology
developers, policy makers in private and public agencies that guide
infrastructure updates and oversee equity policy development, technology
developers that create advanced warnings systems to improve non-motorized user
Decision Makers: State DOT and
local Chief Engineers
Committees: The AASHTO
Standing Committee on Highway Traffic Safety, Standing Committee on Highways
and the Subcommittee on Design could facilitate the implementation of the
research results along with Joint Technical Committee on Non-Motorized
Early Adopters: Early adopters could
be state and local agencies with an AV presence in their jurisdiction or
identified as testing grounds, such as the City of Pittsburgh, M-City, San
Diego, Dallas-Fort Worth and Phoenix.
This research was identified by the TRB Standing Committee on Pedestrians (AFN 10) and is highly recommended as AV technology is rapidly evolving and is already in our roadway corridors. Understanding non-motorized user behavior changes and interactions with AV will help researchers and public agencies identify the appropriate tools to provide a communication bridge between computers and humans as well as ensure communication applications remain consistent as the technology progresses. Bridging this communication gap and creating predictable behavior can increase safety and equity for non-motorized users.
|Sponsoring Committee:||ANF10, Pedestrians
|Research Period:||12 - 24 months|
|RNS Developer:||Amy Rosepiler and Ray Delahanty|
|Index Terms:||Behavior, Human factors, Autonomous vehicles, Communication, Pedestrian vehicle interface, Pedestrian safety, Pedestrians, |
Pedestrians and Bicyclists
Safety and Human Factors
Vehicles and Equipment