Advanced Driver Assistance Systems (ADAS; Level 2 vehicle automation) are increasingly common in vehicles. Although these systems are designed to enhance driving comfort and safety, their capabilities and limitations are often misjudged, potentially creating new traffic risks. While drivers’ misunderstanding and overreliance on ADAS have been implicated in several high-profile accidents, insufficient empirical research has examined how such misconceptions relate to unsafe driving behaviors. In a pre-registered survey of 130 Tesla Autopilot users in China, we measured users’ underestimation and overestimation of Autopilot’s capabilities and their operational knowledge regarding the use of Autopilot’s key functions. Participants were more likely to overestimate than underestimate Autopilot’s capabilities, with mean accuracies of about 40 % and 70 %, respectively. Neither misconception was associated with self-reported engagement in non-driving-related activities (e.g., eating and sleeping) while using ADAS. Participants did not show sufficient operational knowledge, with mean accuracy below 75 %. Operational knowledge was positively correlated with intentions to engage in common non-driving-related activities such as eating (r = 0.25, p = 0.004), but not with intentions to sleep (r = − 0.10, p = 0.235). In addition, their propensity to trust machines (a trait-like tendency) was positively correlated with intentions to engage in both common non-driving-related activities (r = 0.28, p = 0.002) and sleeping (r = 0.30, p < 0.001). We discuss the theoretical and practical implications of these findings, particularly in relation to driver education and training that foster appropriate understanding and use of ADAS.
扫码关注我们
求助内容:
应助结果提醒方式:
