Optimizing the remote operation of autonomous driving: The Role of Onboard and Environmental Camera Perspectives

Abstract

Autonomous vehicles are becoming more prevalent on the roads. However, when the automation fails, a remote operator is needed to guide the vehicle, who must quickly assess the situation and make a decision. This research investigated the role of camera perspective on the performance of remote operators of autonomous vehicles. To test this, a between-subjects design was used in which participants were randomly assigned to one of three conditions (only vehicle perspective, only external camera perspective, or a combined perspective) and shown three different simulated driving scenarios in which autonomous vehicles need guidance. After each scenario, their situational awareness, cognitive workload, and decision-making were measured. The results showed that the combined perspective and the vehicle-only perspective provided better situational awareness compared to the external-only perspective. In addition, the mental workload did not increase with the added camera perspective, which suggests that remote operators may be able to monitor multiple video feeds without being overwhelmed. In addition, the user experience was rated highest in the combined perspective and the vehicle-only perspective. A possible explanation for these results could be that the vehicle-only perspective is more familiar as it is similar to the normal driver's view. However, this study also has some limitations, namely that due to the online set-up it was not possible to capture all relevant aspects that are part of the remote operation tasks and interaction with the system. Therefore, future research should focus more on testing long-term exposure to the system and simulating a more realistic remote control interaction.