Nature
This weekend, almost a million people will pour into London's Notting Hill district for its annual carnival, Europe's largest street party. The authorities will treat them like ants in the hope of making their experience more pleasant.
Researchers have adapted computer models of insect swarms to work out the carnival parade route that is most likely to alleviate the crowding that can mar the event.
In the past, the parade has taken a circular route. This traps people in a small area and intensifies the crush. Two years ago, the Greater London Authority asked geographer Michael Batty of University College London and his colleagues to explore six alternative routes.
Last year, the team monitored the entry points into the carnival area. They combined this with information from transport authorities, first-aid providers and aerial police photos to work out how many people were going where.
The researchers then created a virtual Notting Hill. About 15,000 computerized pedestrians roamed simulated streets to find the quickest routes and prime attractions. People followed each other to entertainment hotspots, but were repelled by dense crowds.
This allowed the team to identify the potential crush spots and to predict the effects of different parade routes, blocked streets or closed underground stations.
The Notting Hill model predicted that an L-shaped parade route would reduce crowding most effectively. This foundered on the carnival organizers' desire to keep the route as circular as possible, and the wish of some local councils to keep the parade out of their area.
The compromise route is U-shaped - an interim solution, says Lee Jasper, a senior policy advisor to the Mayor of London. "We're taking an evolutionary approach," he says, to reconciling political and public safety concerns.
"The key thing is that we've now got several objective standards by which to make decisions, whereas in the past we've just had common sense and experience," says Jasper.
Mass movement
This type of simulation has been used to explain ants' behaviour, whereby many individuals following simple rules can produce complex and orderly transport networks without any outside control. "It's characteristic of how insects are attracted to food," says Batty.
Treating carnival-goers as individuals might work well for people heading into Notting Hill, comments computer scientist Jon Kerridge of Napier University in Edinburgh UK. But such simulations might not reflect the crush of the carnival, where freedom of movement is limited.
"People in highly congested situations behave like a fluid - they move as one," says Kerridge, who studies pedestrian movement. Often, for example, groups hold hands to stay in touch. Simulating individuals' movement might miss the effects of these behaviours on crowding, he says.
"Our model doesn't show the dynamics of crushing on a fine scale," admits Jake Desyllas of Intelligent Space, a London-based consultancy firm that collaborated in the project. "But it shows the key factors that influence crowds on an urban scale." The research team will be moving among the crowds at this year's carnival to collect information to refine their simulation.
© Nature News Service / Macmillan Magazines Ltd 2002
|
