Publication

Despite advancements in automated driving technology, human supervision remains essential in the transport system. Remote operation can support monitoring, assisting, and driving (semi-) automated vehicles from a distance when necessary. However, passive human monitoring and occasional intervention pose challenges, as, for example, critical decision-making can be impaired by the human operator’s state. Decades of traffic safety research have led to the development of Driver Monitoring Systems (DMS) and Cabin Monitoring Systems (CMS). These systems can detect risky driver states and behaviors, such as distraction, fatigue, and engagement in non-driving tasks.

This pre-study aimed to explore the potential of applying existing DMS and CMS technologies for remote operators of automated vehicles. It also sought to identify gaps in these systems and explore improvements to better support remote operators, enhancing traffic safety in future applications of remotely operated automated vehicles.

Four workshops were carried out during the pre-study. Workshop 1 reviewed DMS and CMS technologies, highlighting features like distraction and drowsiness detection, and body posture tracking. It also discussed the need for ergonomic adjustments for remote operators. Workshop 2 focused on the roles of the remote operator as well as required training and communication schemes. Workshop 3 analyzed gaps in current DMS and CMS technologies and Workshop 4 consolidated the findings from the workshops and outlined future needs for Remote operator Monitoring Systems (RMS) as well as plans for further research and development.

The pre-study concluded that DMS and CMS technologies can be adapted for remote operators of automated vehicles, detecting critical behaviors like fatigue and distraction. This adaptation could also allow real-time behavior monitoring and data collection for improved operator performance, mitigating risks from, for example fatigue and disengagement. These technologies could also provide data-driven assessments for improving the remote operator work environment and managing incident liability. However, current DMS and CMS technologies cannot distinguish between negative and positive behaviors, such as deliberate strategies to maintain focus. This limitation highlights the need for “intelligent” assessment of operator behaviors where contextual understanding is needed.

The findings from the pre-study will be disseminated within SAFER’s new Remote Operations working group, with industry, academia, and research representatives. Authorities are also invited. The working group objectives are to network, share knowledge, and initiate new projects, for example a pilot study to develop a functional RMS concept and to address traffic safety challenges associated with remote operations of automated vehicles.