The Research of Multilevel Takeover Alert Information Design for Highly Automated Driving Vehicles

摘要

The highly automated driving vehicles relief a certain amount of driving loads but also bring problems on how to get drivers' attention back to control when it is necessary. This paper proposed a design concept of multilevel alert takeover information and developed a simulated driving system, and the investigated questionnaires were used to evaluate the takeover performance and user experience. Results show that both levels and channels affect alert efficiency. Multilevel alerts are not able to significantly enhance driving takeover performance but effectively improve user experience. Auditory alert information enables drivers to perceive the risks more quickly, while visual alert information assists drivers to learn about the risk degree more efficiently. It is suggested to add personal experience to motivate the takeover ambitious in further study.