A supportive situation awareness model for human-autonomy teaming in collaborative driving

Abstract

ABSTRACTDriving has become a collaborative activity and a form of human-autonomy teaming (HAT) with the addition of autonomy to the advanced driver assistance system (ADAS), which makes situational decisions and sensible actions (e.g., autopilot and collision avoidance). However, it has been identified that in many fatal road accidents involving collaborative driving, over-reliance on the ADAS becomes the primary factor. To overcome this issue, the underlying situation awareness (SA) concept is investigated to identify an appropriate SA model for collaborative driving that could impact the intelligent agent’s design in an HAT context. The formalization of existing SA model characteristics is defined and compared with those in collaborative driving. As a result, existing SA models are inadequate for explaining collaborative driving. Therefore, a new supportive SA (SSA) model is proposed. Based on the nature of this new model, applying transparency during SA development of the ADAS is suggested as a mechanism to comprehend ADAS behaviours. The proposed SA model is a significant expansion of multiple-agent SA models, and a transparent-based system can be a future direction of ADAS development to calibrate drivers’ trust.