Enhancing Participants’ Experience in Fuzzy Control-based Autonomous Racing Car Hackathon through Developing Driving Log Visualization

Abstract

A hackathon at the 40th Fuzzy System Symposium focused on Virtual Motorsport Lab’s hands-on program using autonomous racing car simulators. We developed an autonomous driving function based on fuzzy control and a driving log visualization for this hackathon to support participants in their program development. Most participants, however, concentrated on editing the target trajectory, while only a few explored other aspects of their program. To enhance the development experience, broadening the range of areas participants can engage with, regardless of their prior knowledge, is essential. This study explores two solutions to this issue. The first solution is to enhance the pre-hackathon training by including an exercise on programming the driving log visualization, encouraging participants to focus on and understand aspects beyond the trajectory setting. The second solution involves expanding the information display by incorporating elements into the driving log view that visualize the fuzzy controller’s behavior. This will provide participants with more insights for tuning the controller. We conducted experimental investigations to evaluate the effectiveness of these approaches.