During the annual Hajj pilgrimage over two million Sunni Muslims travel along a proscribed route in and around the Saudi city of Mecca. Tragically, this mass-movement of people sometimes triggers crowd disasters, particularly at places where the route narrows such the Jamarat Bridge, where pilgrims perform the ritual stoning of the Devil.

Dirk Helbing and Anders Johansson of the Dresden University of Technology analysed video images of the January 2006 disaster at the Jamarat Bridge, in which over 300 pilgrims died. They used a computer algorithm that determined the velocity and position of individuals in a large crowd moving towards the bridge. The data covered the 45 minute period up to and including the disaster and revealed three distinct phases of motion.

During the initial phase the pilgrims made steady progress towards the bridge, but the rate of flow decreased as the crowd became denser. Then the crowd suddenly underwent an abrupt transition to a second phase of “stop-and-go” motion that propagated like waves along the direction of travel.

These waves persisted for about 20 minutes as movement towards the bridge slowed and crowd density continued to increase until the crowd made a sudden transition to the third and very dangerous phase. At this point clusters of people began moving randomly in all possible directions in a phenomenon described by Helbing and Johansson as “crowd turbulence”

The researchers believe that turbulence may have been brought on by individuals panicking and pushing in all directions to increase their personal space. This caused violent pressure waves to surge through the crowd, tossing individuals several metres, tearing off clothing and ultimately leading to the trampling of hundreds of pilgrims.

The study revealed that death and serious injury began about ten minutes after the onset of turbulence and 30 minutes after the onset of stop-and-go flow. Helbing and Johansson are hopeful these characteristics could be detected automatically using video systems, giving authorities enough time to implement crowd control measures.

In collaboration with Habib Zein Al-Abideen of Saudi Arabia’s Ministry of Municipal and Rural Affairs, Helbing and Johansson have made a series of recommendations, which have already been implemented in a redevelopment of the Jamarat Bridge and its surroundings. In addition to automated crowd monitoring, the bridge and its approaches have been redesigned to improve crowd flow and minimize delays that could lead to over-crowding. There were no major incidents at the bridge during this year’s Hajj.

This is not the first time that Helbing has studied crowd dynamics and he has identified “shock waves” that travel through dense crowds. However, the much more dangerous crowd turbulence seen at Jamarat Bridge is a phenomenon that has not been studied before.