Behavioral Reactions and Adaptations of Manual Drivers in Mixed Traffic on the Highway – A Longitudinal Driving Simulator Study

• Applicants: Dr. Vanessa Stange; Professor Dr. Mark Vollrath
• Subject Area: Human Factors, Ergonomics, Human-Machine Systems ; Traffic and Transport Systems, Intelligent and Automated Traffic
• Term: since 2023
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 514331715

Today, vehicles controlled by partial automation (Level 2, SAE International, 2021) are travelling on the road alongside vehicles manually driven by human drivers. Recently, the manufacturer Mercedes-Benz started selling an automated Level 3 (L3) vehicle. As a result, "mixed traffic" is emerging in which manually driven drivers will share the highway with these L3 vehicles and increasingly interact with them. Since these L3 systems are expected to be introduced in the premium segment first, it is likely that human drivers will continue to be in the majority for a long time. In mixed traffic, manual drivers face the challenge of interacting not only with other human drivers, but with L3 vehicles in certain driving situations. These vehicles always comply with the rules. In contrast, among manual drivers there is probably only a small fraction of drivers, e.g. novice drivers, who show equally rule-compliant driving behavior. Upon the introduction of L3 vehicles, manual drivers will initially rarely encounter this rule-compliant driving behavior, just as in all-manual traffic today. However, as the penetration rate of L3 vehicles increases, it can be assumed that manual drivers will encounter this rule-compliant driving behavior more and more often. To date, there is limited empirical evidence on mixed traffic interactions on the highway. These studies primarily examine single, brief interactions in initial contact with L3 vehicles on the highway. Beyond that, there are practically no results available so far. The goal of this research project is a longitudinal investigation of the driving behavior of manual drivers in mixed traffic on the highway in a driving simulator. Two learning processes will be investigated. On the one hand, it will be investigated whether manual drivers adapt their existing schemes and scripts, and thus their behavior in the presence of these vehicles, accordingly in subsequent driving, so that potentially dangerous behavior of manual drivers in interaction with L3 vehicles is avoided in the longer term. The second learning process relates to the potential role model function of L3 vehicles for manual drivers with respect to rule-compliant driving behavior even when driving without L3 vehicles immediately present. If this learning takes place, this could represent an indirect contribution of automation to higher traffic safety and better traffic flow. As a possible factor promoting learning, but also possibly reinforcing negative reactions, the external labeling of the automated driving mode is included in the investigation. Thus, this research project contributes to closing this research gap, which has also been identified by different authors in the current literature and described as very relevant for future research.

• DFG Programme: Research Grants