Virtual reality software developed at Middlesex can stop an air emergency turning into a disaster
You're an emergency responder at a major European airport. Smoke is coming from a plane, but you can't see which part of the aircraft is on fire. Then a report comes in on your earpiece of an explosion. Meanwhile, other flights are stacking over the airport, waiting to land. What do you do next? Situations like this are extremely rare, but airport staff must be prepared for them. The problem is that training exercises on the ground don't take place often enough to ensure everyone is up to speed. People leave, new staff arrive and procedures change. But the bespoke "airport in a box" software developed by William Wong and his team allows clients like Keflavik airport in Iceland to simulate a critical incident in real time.
Funded by the European Commission FP7, the 12-partner CRISIS project consortium took a core product developed by the Dutch partner E-Semble, XVR, and adapted and extended its functionality to better train the way people make decisions during emergencies. To do so, the team also drew on techniques used in video games and horror movies to create conditions of surprise and startle.
The challenge was to create bespoke virtual reality environments that reacted to users' actions just as VR games do - but much more cheaply, and with the ability to configure a scenario in hours or days rather than months. Staff adopt an avatar such as a firefighter or a triage medic and communicate through headsets and walkie-talkies as the simulation unfolds. Sounds and bright lights convey the enormity of the event, which can be programmed to take place at night or during the day, or even during extreme weather. "We've moved away totally from philosophies where you tell people what's going on," says Professor Wong. "The scenario unfolds before your eyes, requiring the trainee to determine what is happening." Sometimes helicopter noise will prevent an avatar from hearing the instructions from their headset, or their actions inadvertently trigger another accident.
The team talked extensively to experts from Lisbon and Reykjavik airports and the British Transport Police, who were involved in handling real emergencies when they designed the software, and identified the drawbacks of previous crisis training. One was resisting pressure from above when circumstances call for it; another was dealing with the startle reaction that sometimes slows down a response. Because scenarios can be played back and analysed, staff can practise handling situations that "throw" them.
Professor Wong's work on the startle reaction and experiential cues has attracted a great deal of interest from both the academic community and organisations involved in crisis management. He and his team have presented papers about CRISIS at the 11th International Conference on Naturalistic Decision Making and the 56th Annual Meeting of the Human Factors and Ergonomics Society. The work goes on to improve crisis simulations. "There's a lot more that we want to do," says Professor Wong. "For one, we would like to build a simulation environment that can intelligently and autonomously change its state, as well as in response to various end-user actions."