Long-term behavioural adaptation and learning curve of humans interacting with AVs : Deliverable 1.3 in the EC ITN project SHAPE-IT

Report, 2023

Research on the long-term behavioural effects of automated vehicle (AV) use is very important for a correct assessment of AV’s benefits and risks. Thus to ensure the safe and efficient implementation of automated driving technology, researchers need to understand the concept of long-term behavioural adaptation. A combination of research methods is called for to meet this goal, such as human factors research related to AVs—especially their long-term effects. Human factors research brings many challenges in terms of effort and methodological requirements. The SHAPE-IT project considered an approach that brings this goal closer to reality. This deliverable has two aims. One is to assess the development of drivers’ situational awareness (SA) during automated driving over time. In particular, when drivers receive a Hand-Over Notice (HON), their Take-Over Time (TOT) is influenced by factors known to play a role in their decision-making process, which is a part of their SA. Thus, changes in TOT may indicate a change in SA. Long-term behaviour changes can include safe behaviours as well as risky behaviours - such as performing Non-Driving Related Tasks (NDRT) and misusing the system (whether intended or unintended). The second aim is to assess the development of pedestrians’ SA of AVs over time. Each aim is addressed by a project. In one example, pedestrians' decisions whether to yield to an AV or cross the street first, and the factors influencing their decision-making process, are examined. The drivers’ and pedestrians’ behavioural adaptation is the focus of our research, with behavioural adaptation described as a learning process (Forster et al., 2019) over time.

This deliverable aims to provide the reader with a meticulous discussion assessing and explaining the behavioural adaptations and changes due to long-term human-automated vehicle interactions (HAVI) or repeated AV exposure. The insights from the project have an impact on human factors and ergonomic engineering, research, and development in the context of automated driving. Both projects (driver-AV interactions and pedestrian-AV interactions, over repeated exposure) are discussed in the following sections.