Final Project Report (Swedish)
This project has studied how virtual reality (VR) can be used to study the communication between self-driving vehicles and cyclists from the cyclist's perspective. In the project, a bicycle simulator for implementation into VR has been developed, and an experiment was carried out to test how studies that focus on future vehicle communication can be carried out, and how reliable and realistic the results are.
The bicycle simulator has been developed in the game engine Unity and a virtual environment has been built up that represents a 500-meter-long stretch of road in an industrial area in Lund, Sweden. In the experiment, 24 participants experienced several scenarios where traffic has been simulated based on observations made at the site with drones. The test subjects cycled through two scenarios where they interacted with either electric or traditional diesel trucks, and two validation scenarios where they experienced the same traffic as an observed cyclist did. After the experiment, the test subjects filled in questionnaires about how they experienced the sound from the trucks, how they experienced the simulation in its entirety, as well as specific questions about simulator sickness.
The results from the experiment make it possible to analyze both how the electrification of future vehicles may affect how cyclists interact with trucks and how realistically the participants cycled within the simulation compared to observed cyclists. The results show that the participants were unsure of how big and how close the electric trucks were, which made them more careful. But the results also indicate that the test subjects generally biked faster and were more aggressive in the simulation compared to observed cyclists in the real environment. A relatively large proportion of the test subjects also reported some simulator sickness and two people ended the experiment early due to nausea.
The project has achieved its goals and shown that it is relatively quick and easy to test vehicle communication in VR. Future research should focus on how this type of studies can be implemented in the development process of new technology and to validate that cyclists not only ride realistically but that their assessment of the new technology also agrees with real tests normally happening later in the development process.