Cycling in automated traffic

Scenarios and test criteria

Abstract (2022)

Authors

S.H. Berge Transport and Planning - CEG
J.C.F. de Winter Human-Robot Interaction - Mechanical, Maritime and Materials Engineering
Marjan Hagenzieker Transport and Planning - CEG
Research Group
Transport and Planning (CEG) (TU Delft)
More Info
expand_more

Published Date

2022

Language

English

Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Faculty

Civil Engineering & Geosciences

Department

Transport and Planning

Research Group

Transport and Planning

Abstract

Background: In the future, automated vehicles (AVs) must interpret and adapt to the subtleness and complexity of urban traffic. Interaction with vulnerable road users (VRUs) like pedestrians and cyclists in complex, urban traffic environments is still a major challenge for AVs. While traffic scenarios with pedestrians and AVs are well studied, the field of AV-cyclist interaction is still in early development. With differences in speeds, eye-gaze behaviour, and movement patterns compared to pedestrians, it is crucial to target cyclists as a specific road user group in AV-VRU research. The objectives of this study are to develop realistic scenarios, test criteria, and guidelines for studying AV-cyclist interaction.

Method: We are applying a mixed-method approach in this study: A systematic literature review of scenario-specific studies on cyclists and motorised vehicles along with in-depth analyses of cycling accidents and safety-critical situations, focus group interviews with cycling safety and human factors experts, and real-life testing with cyclists.

Results: The study is ongoing, and results are expected Third Quarter 2022. From the systematic literature review, we expect to derive the most common cyclist interactive scenarios from cyclist research as well as the relevant situational factors of these scenarios. To evaluate the scenarios and to identify additional technological factors relevant to AVs, the scenarios will be assessed by cycling safety and human factors experts in focus group interviews. The interviews are expected to result in a selection of scenarios and a list of test criteria for each scenario. Validation of the scenarios and test criteria will be evaluated using an instrumented bicycle in real-life traffic.

Conclusion: To accommodate the need for more knowledge of AV-cyclist interaction, we will provide a collection of traffic scenarios and methodical guidelines for future research. The findings are particularly relevant for studies on AVs, cyclists, and external human-machine interfaces (HMIs), on-bike HMIs, and infrastructural systems. With a rich description of the situational and technological factors involved in the test scenarios, the scenarios can be utilised to model and predict road user behaviour in future automated traffic.