Multimodal interaction strategies for take-over requests in L3 automated driving

L3 automated driving allows the driver to engage in non-driving tasks, but when the system exceeds or is about to exceed the ODD (operational design domain), the system will remind the driver to take over the vehicle in time. In order to explore the impact of different interaction modes (visual, auditory and tactile) and their combined TOR (take-over request) interaction strategies on take-over performance, five different take-over interaction strategies are designed for TOR. Using the driving simulator and platform equipment, the automated driving take-over experiment was carried out with 24 automotive engineers for 120 person-times' experiments. The results show that when the TOR period is 10 seconds, the multimodal interactive strategies can make the driver take over the vehicle successfully. Strategy 3 (visual + auditory + tactile) takes the shortest take-over time (M=3.8 seconds), which is most suitable for unplanned TOR. In strategy 4 (two-level TOR), drivers have the best performance in taking over the vehicle for the planned TOR with a shot take-over time (M= 4.15 seconds).