Anticipated and Adjusted Crowd Movements to the Actual Terrain Situation – MAASTeR

The MAASTeR project (Mouvements de foule Anticipés et Ajustés à la Situation Terrain Réelle – anticipated crowd movements adjusted to the real field situation) proposes the creation of a decision support tool for security forces and public building operators focusing on security and mobility. This project is part of the field “crowd movement management” and offers an integrated decision support tool to improve the security of Olympic and Paralympic Games, in a context of evolving threats and risks.

The management of crowd behavior is a major issue for the safety of public places, especially in the context of international events with a high concentration of population. In operational management situations, various hazards can quickly reach a level of discomfort or even insecurity, such as the increase in people's densities, the saturation of traffic, or the evacuation of people. New tools are needed to streamline decision-making and secure the consequences.
A first challenge is to manage these risky situations, their effects and their resolution, by testing a range of virtual scenarios. These preliminary tests make it possible to evaluate, improve and disseminate procedures: measure densities to detect dangerous zones, forecast the movements of crowds in normal situation and facing an event, position security forces, train personnel in virtual reality. Thus, modeling can be used as a preventive measure for optimization both in the long term, such as urban planning, and in the short term, such as setting up or organizing events in existing places (like the Olympic Games). These preliminary studies must be able to adapt to the evolution of situations, from planning to operational management.

A second challenge is to adjust the simulations to the real field situation, considering its dynamics. Thus, alerts can be raised as soon as dangerous situations are detected, and even before they occur thanks to simulation. The integration of multi-source capture data (video, Bluetooth, Wifi) with a high-performance simulator makes it possible to obtain a synthetic view of the situation, but also to react more quickly during events by predicting their effects. The proposed MAASTeR solution allows an organization or security forces to react to current and observed situations, by anticipating the decisions to take facing an event in order to ultimately prevent accidents.

The proposed technological innovations integrate dynamic situation assessment through video and IoT (Internet of Things / Internet of Things) with a crowd simulator.

Video capture will extract macroscopic data (flows, densities of people). IoT (Bluetooth / Wifi) capture will provide microscopic information (individual trajectories). These two combined data sources will be used to calibrate a crowd simulation, by setting the number of entities (density), their distribution (flow) and the origin / destination matrices.

Fast simulation will provide a synthetic view of the current situation and in the near future (a few minutes). This decision support solution will allow operators to be more agile facing hazards, but also to create models with high added value for planning.