Measuring cyclists' visual attention and interaction with an interactive real-time 3D simulation

In recent years, the safety of Vulnerable Road Users (VRUs) has improved much more slowly than that of motorized road users. A promising solution are Advanced Driver Assistance Systems (ADAS) for cyclists in the context of Vehicle-to-X (V2X) communication. To ensure their effectiveness and safety, these systems have to be tested psychologically without risking the lives of participants. Suitable simulators such as the Virtual Cycling Environment (VCE) can help with this task by letting cyclists interact with simulated networked traffic. So far, such simulators have rarely been used for VRUs in long scenarios with complex traffic. In this thesis, I evaluate the suitability of the improved VCE in an empirical study and I present a method for measuring cyclists visual attention in different traffic densities. Results align with the literature in showing that higher-density traffic requires more attention. The VCE is therefore considered suitable for human-in-the-loop experiments. Acting on feedback from participants can further enhance this suitability, e.g., by improving the way other cars react to the cyclist.vorgelegt von Lukas Stratmann ; angefertigt in der Fachgruppe Distributed Embedded Systems (CCS Labs), Heinz Nixdorf Institut, Universität Paderborn ; Gutachter: Falko Dressler, Ingrid ScharlauTag der Abgabe: 31.05.2019Universität Paderborn, Masterarbeit, 201