Riding behavior-based HMIs for motorcycle safety (A comparative study of alert and real-time feedback systems)

Abstract

Motorcycle assistance systems have been an increasing focus of research in recent years to promote safer riding practices and mitigate motorcycle-related accidents. While there is substantial interest in evaluating the beneficial aspect of these systems in improving driving safety, it is equally essential to clarify their unintended adverse effects on riders' driving behaviors and the overall interaction between riders and such systems. In this study, we designed alert and continuous real-time feedback HMIs that convey information about lateral deviation from the lane center based on our previous research results. Experiments were then conducted on a riding simulator to investigate whether variations in information presentation could alter rider-system interaction and ultimately affect riding performance. Fifteen young motorcyclists (mean age = 22.1 ± 1.87 years) were exposed to dummy and risky scenarios at uncontrolled intersections while interacting with alerting or real-time feedback systems separately. A path model was constructed, and the impacts of HMIs on riders' reaction time, information observation time, weighted workload, and risk of collision were clarified. Cluster analysis was subsequently utilized to categorize participants into confident and methodical riders, predicated upon their riding styles. Independent t-tests revealed that, except for Obs-time, there were no significant differences between confident and methodical riders in RT and TTCmin when using an alert HMI. However, when real-time feedback was displayed, significant differences were found for TTCmin and Obs-time. In general, participants exhibited superior performance with the Alert HMI, characterized by a shorter RT (1 second), longer TTCmin (0.79 seconds), reduced information observation time (0.42 seconds), and lower perceived workload compared to when real-time feedback was used. The authors believe that this research will contribute to the development of more effective and rider-centric HMI designs.