Abstract

In the near future, mixed traffic consisting of manual and autonomous vehicles (AVs) will be common. Questions surrounding how vulnerable road users such as pedestrians in wheelchairs (PWs) will make crossing decisions in these new situations are underexplored. We conducted a remote co-design study with one of the researchers of this work who has the lived experience as a powered wheelchair user and applied inclusive design practices. This allowed us to identify and reflect on interface design ideas that can help PWs make safe crossing decisions at intersections. Through an iterative five-week study, we implemented interfaces that can be placed on the vehicle, on the wheelchair, and on the street infrastructure and evaluated them during the co-design sessions using a VR simulator testbed. Informed by our findings, we discuss design insights for implementing inclusive interfaces to improve interactions between autonomous vehicles and vulnerable road users.

Reference

Abstract

We are interested in the ways pedestrians will interact with autonomous vehicle (AV) in a future AV transportation ecosystem, when nonverbal cues from the driver such as eye movements, hand gestures, etc. are no longer provided. In this work, we examine a subset of this challenge: interaction between pedestrian with reduced mobility (PRM) and AV. This study explores interface designs between AVs and people in a wheelchair to help them interact with AVs by conducting a preliminary design study. We have assessed the data collected from the study using qualitative analysis and presented different findings on AV-PRM interactions. Our findings reflect on the importance of visual interfaces, changes to the wheelchair and the creative use of the street infrastructure.

Reference

Abstract

In the near future, mixed traffic consisting of manual and autonomous vehicles (AVs) will be common. Autonomous vehicles with advanced technology offer opportunities for innovative designs and introduce communication challenges for vulnerable road users such as pedestrians and cyclists. Our goal is to explore the emerging new domain of interaction between different road users and autonomous vehicles in a future AV transportation ecosystem. This led us to conduct the thesis following these two themes: 1) understanding design opportunities for external automotive displays (EADs) of AVs; 2) exploring the design of interactions between vulnerable road users (VRUs) and AVs. In theme 1, our work extends contemporary research into visualizations and related applications for autonomous vehicles. Focusing on external car bodies as a design space we introduce a set of EADs. EADs show visualizations to share context and user-specific information and offer opportunities for interaction between users and AVs. We conducted a design study to explore design concepts for EADs to provide services to different road users: pedestrians, passengers, and drivers of other vehicles. Based on the design study, we prototyped four EADs in virtual reality (VR) to demonstrate the potential of our approach. This exploration contributes to our vision for EADs, a design critique of the prototypes, and a discussion of the possible impact and future usage of external automotive displays. In theme 2, we are interested in the ways pedestrians will interact with autonomous vehicles in the absence of non-verbal cues from the driver (such as eye movements, hand gestures, etc.). Crossing streets in these new situations could be more dangerous for VRUs without a proper communication medium. We examined a subset of this challenge with two groups of pedestrians: interaction between AVs and pedestrians with hearing aids (PHAs), and pedestrians in wheelchairs (PWs). First, we worked with hearing aid users as a preliminary exploration of this research. We conduct a co-design study with a co-designer with hearing impairment who has lived experience of wearing hearing aid enhancements. This study contributes several insights and design recommendations on how potential audio cues can be designed to enhance direct communications between PHAs and AVs. For the second part of our research, we designed interactions between pedestrians in wheelchairs and AVs. From an early exploration of potential interface designs through a design study with interaction designers, we prototyped different interfaces in VR. Then, we evaluated the implemented simulations during a co-design study with a powered wheelchair user following inclusive design practices. We identify and reflect on interface design ideas that can help PWs make safe crossing decisions at intersections and discuss design insights for implementing different inclusive interfaces.

Reference