Mitsubishi Electric System of Forecasting of the Movement of Crowd

2016: Announcement

In August, 2016 the Mitsubishi Electric company announced creation of the first-ever IT system for prediction of the movement of crowd in real time. The new technology should help to avoid accidents during the populous actions.

The system developed in cooperation with the Research center of modern science and technologies (Research Center for Advanced Science and Technology, RCAST) at the University of Tokyo, analyzes data from surveillance cameras in venues of large actions and in real time gives very exact assessment and the forecast of flux density of visitors on an input and an exit.

The Mitsubishi Electric system studies behavior of each person in accumulation of the people and the next to it people while other technologies such estimate the movement of crowd in general

Developers claim that the software is capable to predict the movement of crowd with an accuracy of 80% whereas the indicator of traditional methods of measurement is only 30%.

Studying of behavior of each person in accumulation of the people and the next to it people while other systems such estimate the movement of crowd in general became feature of a new product.

It is supposed that new development will help to increase security and efficiency of regulation of flows of people in places of their big accumulation. Forecasting of places of presence of a huge number of people at real time will allow event managers to take timely adequate measures to unload points of expected jams.

Such system is useful at sports stadiums: seeing, much more everything there are fans after the match, organizers will be able timely to open and close exits, having provided the fastest and at the same time safe escape of people.

In Mitsubishi Electric note that event managers with a large number of visitors often should increase considerably the number of personnel to provide the necessary security level both in the venue of an event, and in the adjacent territory. The existing technologies of assessment of flux density are constructed on use of data on the basis of the last periods, include very difficult calculations and do not differ in adequate accuracy.^[1]

Notes