Boosting Perception of Relevant Road Users Under Adverse Weather Conditions: Evaluation of a Human-Machine Interface for Remote Assistants of Highly Automated Vehicles

Abstract

Remotely operating vehicles enables the effective use of vehicle automation even when fully automated driving is not feasible. A human operator assists the vehicle automation remotely when its capabilities are exceeded. The vehicle’s remote assistant, legally defined as Technical Supervisor (“Technische Aufsicht”), is a prerequisite to operate highly automated vehicles (HAVs, SAE 4) on public roads in Germany. Assessing traffic situations and providing assistance heavily depends on the video stream from the surroundings of the supervised vehicle. However, adverse weather conditions may deteriorate the visibility of relevant road users, decreasing the remote assistant’s performance in assessing and acting on traffic situations that the vehicle’s automation cannot handle independently. This poster presents a human-machine interface (HMI) concept for augmenting the remotely assisted HAV’s video stream displayed at the remote assistant’s workplace. The video stream is overlaid with boundary boxes around relevant road users, boosting salience and improving the remote assistant’s perception. The required data can be generated by conventional HAV sensors. The HMI concept was evaluated in an experimental user study with modulated workload via a secondary task regarding performance, workload, situation awareness, and usability. The study’s results inform the iterative development of a workplace for remotely assisting HAVs.