This project aims to provide information regarding the movement and behaviour of crowds present at large events such as concerts, open markets, etc. Using this information, the organizers of such events will be able to better plan the necessary infrastructure and organize their security services to quickly react in dangerous situations such as panics, fire, etc. We consider using two main sources of data: posts and interpersonal relationships on social networks and images from surveillance cameras.

The data collected from social networks allow having an “a priori” knowledge about the number of participants at an event and at the same time information about their attitude during the event. The social media can therefore serve as very good indicators of the mood and the behaviour of the crowd, thereby preventing violence and enabling the optimal ways to ensure the physical safety of the people.

Camera images allow following in real time the movements of the crowd, identifying conflicting situations and thereby quickly raising alerts in case of a potential danger.

In addition, a particular emphasis in this project is placed on safety and with respect to it, we consider two aspects: the first one is regarding the physical security, allowing a correct estimation and adaptation of security services and a fast reaction in case of need, and the second one is protecting the private sphere of the participants by ensuring anonymity of treated data. In order to provide a good scalability of collected data, the processing will be implemented on the cloud. Contact : Didier Rizzotti


Didier Rizzotti

HES-SO Rektorat
Route de Moutier 14
2800 - Delémont
T +41 58 900 00 00 - F +41 58 900 00 01