University–Industry Partnership for AI-Powered Antiterrorism and Public Security

This project aims to develop an AI-based system capable of identifying potentially threatening behavior in crowded public environments such as airports, metro stations, and transportation hubs. The core assumption is that individuals under extreme stress—such as those preparing for malicious acts—exhibit subtle behavioral anomalies, particularly in their movement patterns.

Using video surveillance data, the system will extract real-time movement-related features (e.g., walking speed, step variation, directional changes) and model these as time-series inputs. Advanced machine learning algorithms, especially LSTM-based neural networks, will learn normal motion behaviors and detect deviations suggesting elevated risk or intent.

We are looking for partners with expertise in:

• Artificial Intelligence and Deep Learning

• Behavioral Science and Human Dynamics

• Computer Vision / Video Analytics

• Public Safety Authorities (e.g., Police, Airport Security)

• Smart Sensor Integration / IoT Surveillance

The expected outcome is a modular and scalable early warning system that can assist security forces by highlighting suspicious activity patterns before harmful actions occur. Ethical design and privacy preservation will be a key consideration throughout the development process.

Related Horizon Europe Topic:

Better protect the EU and its citizens against Crime and Terrorism – Horizon Europe Cluster 3