External Human-Machine Interfaces for Autonomous Vehicles to Pedestrian Communication

This research aims to understand how pedestrians interact with Autonomous Vehicles (AVs). The envisaged outcome is to develop interfaces, guidelines, or recommendations to help pedestrians safely interact with AVs within Singapore’s context.

Findings from socio-behavioural user studies will contribute to the development of an External Human-Machine Interface Simulator System used to evaluate safe pedestrian-AV interactions.

Objectives

The project comprises two main objectives:
The first is to develop recommendations to educate pedestrians on interacting with AVs, and behavioural guidelines for AVs, based on socio-behavioural user studies.

Pedestrian studies and design workshops will help us understand how pedestrians use the road with human drivers under controlled conditions, how they communicate intentions, and which behavioural characteristics can be leveraged. These findings will inform the guidelines and recommendations.

The second objective is to design and implement a software simulator for external vehicular interfaces, behaviours, and design specifications for AVs.

The External Human-Machine Interface (eHMI) simulator conveys potential AV behaviours, including critical events like near-misses that cannot be safely recreated in real life. User studies with the simulator will evaluate the proposed guidelines and designs, informing design specifications for eHMIs for AVs.

Publications

R. Y. H. Sim, P. C. Ng, K. Chen, and J. S. A. Lee, “Pipeline for text-to-image panoramic road scene generation and evaluation,” in Adjunct Proceedings of the 16th International Conference on Automotive User Interfaces and Interactive Vehicular Applications, Automotive UI ’24 Adjunct, (New York, NY, USA), p. 143–148, Association for Computing Machinery, 2024

J. S. Lee, A. Yeo, B. W. Kwok, Z. F. Wong, E. Poh, and R. C. Chan, “Intvrsection: Virtual reality environment for evaluating signalized and unsignalized intersection scenarios,” in 15th International Conference on Automotive User Interfaces and Interactive Vehicular Applications (AutoUI-Adjunct), 2023

A. Yeo, B. W. Kwok, Z.-F. Wong, G.-X. Koh, R. W. Tan, K. Yamin, Y. Liew, D. H. Yan, and J. S. Lee, “Cycling simulation in virtual reality for autonomous vehicle traffic scenarios,” in 2023 IEEE International Conference on Service Operations and Logistics, and Informatics (SOLI), pp. 1–6, 2023

Acknowledgements

This work is funded by:

Research & Development of Methods for a Holistic Evaluation of AV Safety under Tropical Urban Conditions, LTA (UMGC-L010)