Verification of Continuous Recovery Operation Using Teleoperation System for Autonomous Vehicles - Consideration Focusing on Impact of Road Markings

The concept of remote-type autonomous driving was proposed to complement the imperfections of autonomous driving systems, where human operators intervene remotely to ensure proper driving. In a remote-type automatic driving scheme, it is desirable for an operator to monitor multiple vehicles simultaneously to increase work efficiency. To realize this operational structure, this study examines the cognitive challenges faced by operators when multiple recovery requests are simultaneously received from multiple autonomous vehicles. We conducted a user study using a driving simulator by simulating a situation in which recovery requests were continuously received from 10 autonomous vehicles to a single operator. From the results, it is found that the working time increased when there were a series of requests for intervention on factors related to road markings. We then conducted an additional experiment to verify the impact of road markings in a situation requiring recovery. To test this effect in detail, a test similar to the first experiment was conducted in a more realistic scenario involving several traffic participants. Although no statistical differences are identified, the mean work time increases slightly when there are road markings in consecutive situations.