Mob stampedes have killed thousands of people in recent years, but they are usually explained in terms of psychology. Now, European scientists say they can predict and prevent crowd panic by computer simulations using the laws of physics.
The new computer model relies on distances, sizes and velocities instead of emotional states but produces results similar to actual panics, the researchers said in today's issue of the journal Nature.
"We think it works particularly well in panic situations because people don't think about what they should do," said Dirk Helbing, a professor at the Institute for Economics and Traffic at Dresden University of Technology.
The computer models reflect the conditions of a room from which people are trying to escape but cannot use or see exits. The virtual victims appear as particles that reflect an average person's speed, size and desired distance from others.
Under normal circumstances, a crowd exits a theater or stadium in an orderly and coordinated fashion because everyone is moving at a leisurely pace and at an adequate distance from one another, the researchers said.
But when the speed of the individuals increases in a panic, they bump into each other, creating friction and violating personal space. As a result, almost everybody moves less quickly.
Eventually, solid arch-shaped barriers of people clump around the exits and even fewer people can pass to safety. Victims collapse and are trampled, creating further obstacles for the others.
"People want to leave faster, but the result is that they are leaving slower and then the tragedy begins," Helbing said. "The question is what can you do about that?"