An App to Keep Danger at Bay

The Love Parade in 2010 or the Stanley Cup Riots in 2011: events like these show just how dangerous large crowds of people can become, especially when safety personnel don’t have access to real-time monitoring information. Caught unaware, officers on the street can barely take action in critical situations. In order to avoid such an event, police forces in London used free apps for crowd monitoring during the Olympic Games in 2012. The smartphone apps provided by the City of London Police (City Police) and the Westminster City Police (What’s on) sent geolocation data from registered users’ mobile phones to a server, which tracked the movements of thousands of sports-mad visitors around the streets of Central London.

Push-messages inform the user

Using crowd monitoring, the London police was able to follow the masses of people during the Olympic Games, identify safety hazards and directly inform visitors of potential dangers. Geographical coordinates, sent with user consent, were analyzed and displayed on a map. This visualization made it possible to see where and how fast crowds were moving and to evaluate potential risks. Safety officers were informed well ahead of time, allowing them to intervene more effectively. And users were informed about risky situations via push-messages, so they could avoid crowded areas.

The technology was created within the framework of the four-year EU project SOCIONICAL. Development work was performed by a team of students led by Professor Paul Lukowicz in the field of embedded intelligence at the German Research Center for Artificial Intelligence in Kaiserslautern, Germany. Disasters like the 2010 Love Parade in Duisburg, Germany, where 21 visitors were killed when a crush of bodies blocked the event’s entrance/exit, point towards the necessity of this type of app. Users are also convinced of the concept. According to a survey, over 80 per cent of respondents would follow suggestions given by such an app. This is a promising response: As Lukowicz explained, a crowd monitoring app only needs about five percent participation from visitors at an event to effectively reduce danger.