At some point in the future, it might be common to see automated vehicles (AVs) driving next to human drivers on public roads – in so-called mixed traffic. One arising question regarding mixed traffic considers the influence AVs will have on the traffic system. The expected effects of AVs range from a better traffic flow to overall more safety due to AVs rule-compliance.
Another possible impact of AVs on traffic is almost not discussed yet: The influence that AVs might have on human drivers (HDs) through imitation learning. This project closes this gap and investigates possible role model effects from AVs to human drivers.

A simulator experiment was conducted, where the participants are driving in a city environment. The participants stop at a stop sign at a junction, where ten role model cars cross. Following the Social Learning Theory, the participants (observers) are being exposed to the presented behaviour of those role models with the intent to make the observers copy the observation.

To check the hypothesis the models’ behavior is being varied in 15 scenarios. The variation occurs in the distances the model cars are keeping (small, optimal, large) and their driving modes (AC or HD).
The measured variable is the distances the participants are keeping after the observation phase.

The analysis will investigate the influence of the observed distances on the executed distances. Additionally, the second variable (driving modes) allows to explore whether AVs have a different influence on HD’s behaviour than cars driven by other humans.