Self-driving emergency assistance! New DfT research explores passenger needs in nightmare fire, flood and sickness scenarios
The pace of change in self-driving can be dizzying at times. It was only last November that Dr Clare Mutzenich picked up our 2024 Self-Driving Industry Award for Research, for her groundbreaking work on generational trust in AVs for Lacuna Agency.
Fast forward eight months and Lacuna, sadly, is no more, yet Mutzenich is busier than ever – launching a new company, hosting one of the best sessions at MOVE 2025, and conducting more essential research for the UK Department for Transport.
An incredibly detailed 106-page DfT report, published in June, explores multiple emergency scenarios, focusing particularly on the needs of passengers when there’s no driver on-board, as project director Dr Clare explains…
Q: So, Dr Clare, what’s this new DfT study all about then?
“The government has promised that automated vehicles (AVs) will be on the road by 2026 and, as deployment gets closer, there are many questions about how people will interact with AVs in everyday conditions.
“Our study asked, instead, what happens when something goes wrong in a self-driving taxi and there is no driver to help, particularly if users are, for example, older adults, children, or have accessibility issues?
“We considered every one of the nine protected characteristics from the Equality Act 2010, assessing whether characteristics such as age, disability, gender reassignment, marriage and civil partnership, pregnancy and maternity, race, religion or belief, sex, or sexual orientation affected what people could do in emergencies like a fire, flood, or unwell passenger.”
Q: That’s a complicated sample. How did you find the participants?
“We worked with recruiters who reached out to communities, charities and faith groups across Leicester, Nottingham and Birmingham. Everyone was selected based on intersecting characteristics – no one is just ‘disabled’, for example, they may also be male, catholic, black, heterosexual, and so on.
“We aimed for a fully diverse cross-section so we could understand how all users might respond to emergencies. We also removed barriers to participation. Carers could join, venues were accessible, and we avoided scheduling during religious festivals.
“We phoned people in advance to ask how we could make the experience more comfortable; like having fewer people in the room for neurodivergent participants, or scheduling around childcare for mothers. A key focus of this research was ensuring that these seldom-heard voices in transport were included so that inclusive design is baked in, right from the start.”
Q: Amazing. Please tell us more about the methodology…
“We used simulated emergencies. No one was really put into a fire situation! We created six realistic scenarios in virtual reality (VR), including internal issues like an unwell passenger, external interactions like a pedestrian trying to get into your self-driving taxi, and other issues like flooding or being dropped off in the wrong place.
“We tested 91 participants in person. They first viewed the VR scenario while we acted as a virtual moderator, flying around the scene and asking what they would need from the AV system and what they felt they could do.
“Then we took off the headsets and continued the conversation in real life. It’s an exciting, novel method – letting participants experience an immersive emergency in VR with other ‘passengers’, and then immediately reflect on it in an in-person focus group.”
Q: And the key findings were?
“Having no driver to help in an emergency made a big difference, particularly for women, younger and older participants, and those with physical, visual or cognitive impairments. What came across loud and clear was that the responsibility shifts to the user, who might need to speak to emergency services or other passengers.
“Designers will need to ensure that communication is multi-modal and accessible, for example, audio-visual alerts that clearly inform passengers what is happening. Safely exiting the vehicle was a key concern in the fire and flood scenarios. Some participants told us they would need doors or ramps to be activated or would need assistance to get out.
“As we know, disability is not universal – while some users had no issue being dropped further from their destination, for others that would be impossible. Female, non-binary and pregnant women expressed feelings of vulnerability about being alone in a self-driving taxi, especially if a pedestrian tried to enter.
“In that situation, it wasn’t always clear if they should stay in the AV with the doors locked or get out. They wanted the ability to contact a remote operator for support. Race, religion or sexual orientation sometimes made participants feel hesitant about helping or asking for help, in case they were misjudged.
“Even after an emergency, some needed help arranging alternative transport to complete their journey, sometimes preferring not to use self-driving transport again.”
Q: Knowing all that, what would be your call to action?
“The barriers faced by passengers in an emergency in a self-driving taxi aren’t always practical – they can be emotional, sensory, physical and cognitive. Many reactions are shaped by perceptions of safety linked to protected characteristics.
“The absence of a driver presents an opportunity to reimagine how support is offered and could even represent an enhanced, more inclusive service, such as multilingual options or better accessibility features.
“As the technology develops we could tailor transport assistance based on users’ individual needs, like voice-guided systems and inclusive interfaces. But, most importantly, this study shows that we can’t design transport services for people if we don’t ask them what they need.
“I learnt something new every day of testing, as participants shared their lived experience of what works and what doesn’t. For AVs to succeed, we need an equitable approach to design, otherwise the people we most hope will benefit may decide the risk isn’t worth it.”
Specialising in behavioural science, policy and human factors, with a focus on trust, safety and accessibility, Dr Clare’s new company, Anthrometric, is there to support organisations in designing services that work for everyone.
The full DfT report “User requirements to enable passengers of automated passenger services to perform journey tasks during emergencies” is available for download here
