Acceptability and perceived safety of automated vehicle driving styles in mixed-autonomy traffic: Insights from driver-AV interaction

The seamless integration of automated vehicles (AVs) into mixed-autonomy traffic necessitates AV behaviour that human drivers can accurately anticipate and accept. While extensive research examines human reactions to AVs, a critical gap persists in understanding human drivers' explicit expectations of AV behaviour during complex road interactions. Addressing this gap, this study investigates UK drivers' expectations of AV tactical decision-making, exploring how right-of-way priority, conflict types, and AV driving style collectively influence behaviour acceptability and perceived safety. A total of 103 UK drivers participated in a simulation-based video survey depicting unsignalised conflicts, varying across three right-of-way conditions (AV priority, human-driven vehicle priority, unclear priority), two conflict types (crossing, weaving) and two driving styles (aggressive, defensive). Generalised linear mixed models (GLMM) were fitted to the 3 × 2 × 2 within-participants design, with demographic factors included as covariates. The results reveal clear context-dependent preferences: drivers expect AVs to assert priority when holding the right of way and exhibit defensive behaviour when human drivers have priority. In ambiguous situations, defensive driving is generally preferred, though aggressive behaviour is acceptable in low-speed crossing conflicts. Moreover, annual driving mileage influences these preferences in a potential non-linear way. These findings advocate for context-adaptive and human-centred AV planning algorithms that integrate formal rules with human expectations, moving beyond universal “always-yield” policies. This contributes to enhancing AV predictability, minimising conflicts, and fostering public acceptance in mixed-autonomy traffic. • Examined drivers' expectations of automated vehicles'(AVs) tactical decisions. • Drivers prefer an aggressive driving style when AVs have the right of way. • Defensive driving is preferred when drivers have priority or priority is unclear. • Findings guide context-adaptive AVs planning and clear intention communication.