This project has received funding from the European Union’s Seventh Framework Programme for research, technological development and demonstration under grant agreement No. 607626.



SAFECITI is a simulation platform that addresses the requirements of the civil security intelligence community for the specific problem of crowd safety and anticipated threat control in cities. The platform is designed for training purpose and has been developed following the feedback from the Spanish Police. It is not only a game in the context of serious game technology, but an advanced simulator that aims to predict the complex behaviour of the crowd in response to police actions and to provide the analysts with new technology to improve the tactics and operations that are used in real situations.

The product includes relevant technical achievements that bring important advantages to other existing tools in the market.



SAFECITI includes different interfaces and different functionalities for each profile of users (instructor or trainee). Some advantages over other products are that SAFECITI allows several trainees in the same simulation and provide real time radio communication between trainees over the network. The exercises can be created with a high detailed crowd and Police forces deployment, including crowd barricades and Police barriers.



Unlike other tools, complex behaviours (and behaviour propagation) are implemented in SAFECITI. The command set provided is richer than others and customizable.

To create a realistic behaviour of protesting civilians, we needed to represent the overall tension within the crowd and make it spread between the protesters. We also needed to make the civilians perceive the tension around them generated by other civilians, police units and weapons. And finally civilians had to react to the tension according to their behavioural profile. To achieve this, we built a dedicated behavioural Artificial Intelligence (AI).

Adding the behavioural AI really makes the simulation come alive. Civilians act differently according to their profiles and their levels of aggressiveness increase or decrease in reaction to the actions of the police. The crowd behaves heterogeneously and the overall emergent behaviour, is much more plausible than what can be seen in existing commercial products dedicated to Crowd Control.

Furthermore, the impact of the behavioural AI on the overall performance is negligible and even improves performance in some cases. It allows us to fully simulate behaviours of around 10.000 people in a crowd in real time, much more than other products.



Trainees can control the Police units and equipment assigned by the Instructor and give their units high-level commands. The agents are deployed in formations that are edited by the users. The equipment features can be edited as well. This exceptional functionality makes SAFECITI highly customizable to different customers, which is one of its most valuable capabilities for the end users.



The combination of GIS data harmonization, procedural modelling of urban environments and export to a crowd simulation platform in one tool is unique. The system is able to replicate large urban environments of worldwide real cities and represent the scenarios in 2D and 3D views in real time.



As a distinctive feature, the system can replay recorded simulations for their evaluation and analyses. The system can recommend actions during an exercise based on previous similar situations through the use of deep learning functionality according to scores.