Reproduce “Cyber Racing Coach: A Haptic Shared Control Framework for Teaching Advanced Driving Skills” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Congkai Shen, Siyuan Yu, Yifan Weng, Haoran Ma, Chen Li, Hiroshi Yasuda, James Dallas, Michael Thompson and 2 more
Researchers at the University of Michigan and Toyota Research Institute developed "Cyber Racing Coach," a haptic shared control system with a fading assistance scheme designed to teach advanced driving skills. This framework, which integrates a high-performance autonomous racing agent, enabled human drivers to achieve superior racing performance, faster lap times, and greater consistency compared to self-learning or constant full assistance.
Key insights
- Adaptive, fading haptic shared control is critical for effective long-term skill acquisition in complex motor tasks, as constant, full assistance provides no additional benefit over self-learning and can hinder independent skill development.
- High-performance autonomous systems can effectively serve as 'coaches' for transferring advanced driving skills to humans through haptic interaction, provided the assistance is dynamically adjusted.
- The design of human-automation collaborative systems should incorporate intelligent 'scaffolding' mechanisms that adapt assistance to the human's learning progression to foster genuine skill development and prevent reliance on automation.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.