Researchers at Loughborough University used immersive virtual reality (VR) to examine how people respond in emergency situations inside self-driving taxis, an approach designed to inform safer and more inclusive autonomous vehicle design. The project, funded by the United Kingdom’s Department for Transport and the Centre for Connected and Autonomous Vehicles, placed 91 participants wearing VR headsets inside a simulated autonomous taxi during a range of crisis scenarios, including a passenger falling ill, a fire inside the vehicle, flooding along the route, a collision with another vehicle, and a pedestrian attempting to open the taxi door. The study aimed to observe authentic emotional and behavioral reactions in high-risk situations that could not be safely replicated in real life.
Participants spanned a broad range of characteristics protected under the United Kingdom’s Equality Act 2010, including age, disability, and gender, with children aged 8–17 included alongside adults. Researchers asked them what actions they would take, what information or features they would need to manage each emergency, and how safe they felt during the experiences. The findings underlined that barriers to effective emergency responses extend beyond practical limitations to include psychological factors such as perceived safety, confidence, and clarity of guidance when no human driver is present.
Project Director Dr. Clare Mutzenich and colleagues concluded that autonomous systems will need to assume a more active role in supporting passengers during uncertainty and emergencies rather than relying on a human driver. The inclusive design perspective, which incorporated voices often overlooked in transport planning, points toward future self-driving taxis that are not only technically able but genuinely accessible and trustworthy for diverse user groups.
Alongside the VR work, the team contributed to a related driving simulator study with University College London, funded by the same bodies, exploring which non-driving tasks passengers can safely perform while an autonomous vehicle is in control and how quickly and safely they can resume control when prompted.