Implicit intention communication as a design opportunity for automated vehicles: Understanding drivers’ interpretation of vehicle trajectory at narrow passages

Abstract

To ensure road safety in mixed traffic, automated vehicles (AVs) must be equipped with distinct and easy-to-understand communication strategies. For this, the communication design of AVs might be oriented toward implicit communication between manually driven vehicles. This research focused on how drivers interpret observable vehicle behavior (vehicle trajectory) to predict vehicle intentions and planned maneuvers. In addition, the processing and interpretation of contradictory cues were examined as a possible cause for misunderstandings and failed coordination. Two video-based experiments investigated the presumed intention (yielding or insisting on priority), intention recognition time, distinctiveness, and cooperativeness of a vehicle’s implicit communication at narrow road passages. For this, two vehicles approached a bottleneck from opposite sides, and only one vehicle could pass at a time. The driving behavior of the oncoming vehicle consisted of longitudinal (decelerating, stopping, accelerating, maintaining speed) and lateral movements (driving to the center, driving to the edge of the road, no lateral deviation) at different timings (early, late). The results indicated advantages of lateral vehicle movements for distinct intention communication. Lateral movements were interpreted the fastest, perceived more distinct, and were more decisive for the presumed intention than longitudinal movements. Early communication was preferred by drivers but increased the intention recognition time, especially for longitudinal movements. However, early lateral communication was interpreted as fast as late longitudinal communication. Furthermore, lateral and longitudinal vehicle movements within a driving pattern that contradicted each other led to ambiguous perceptions. For AVs, early distinct communication via lateral vehicle movement is recommended. Minimizing ambiguity in the interpretation of communication cues might maximize the probability of the safest response from human drivers and contribute to accident prevention in mixed traffic.