Search

Highly automated vehicles in urban areas will adopt a safe and cautious driving style, which is likely to result in frequent stops and yields while negotiating interactions in complex traffic environments. Users may not expect, comprehend or accept such behaviors from their vehicle. Considering that most daily trips are completed under time pressure, drivers may feel frustrated and decide to take over manual control in an attempt to save time. The current study aimed at investigating the impact of time pressure while driving a highly automated vehicle in urban areas in terms of anger and frustration. Forty participants drove a highly automated vehicle with or without time pressure in a driving simulator. The driving scenario included rural roads with light traffic and urban roads with traffic congestion and driving impediments. Data on mood, attitude and driver behavior were collected. Results indicate that urban automated driving in congested conditions deteriorated participants’ mood, independently of time pressure. Additionally, the number of takeovers on driver’s initiative was higher under time pressure. This study demonstrates that driving a highly automated vehicle in urban areas may have adverse effects for user’s emotional state. As long as drivers are uncomfortable with the vehicle behavior, they may decide to takeover in complex situations. Therefore, it is important to consider the emotional dimension of driving highly automated vehicles. Otherwise, expected benefits of automation for traffic congestion and safety may be compromised.

The complexity of urban driving environment may create frequent interruptions in the course of automated vehicles, which may have adverse effects on driver's mood, and even result in driver-initiated takeovers. A possible way to counteract this is to help drivers reappraise the situation by providing additional information when they encounter interruptions in vehicle's progression. This study aimed at assessing the effect of an HMI enhanced with contextual information on drivers' mood, attitude and behavior compared to a basic HMI. Participants drove the experimental scenario in a driving simulator, once with and once without receiving contextual information. Results are discussed, potential bias is identified, and perspectives are evoked to improve the proposed method.