As part of the #REG4IA-TER initiative (the Regional AI Hub for Territory Resilience and Security led by the Tuscany Region) the "Smart Presence" subproject aims to innovate emergency management in complex environments. The project develops multimodal and generative AI models to monitor personnel flows inside and outside critical infrastructures, with a specific focus on the Cisanello Hospital area. The primary goal is to provide decision-makers with advanced predictive and descriptive tools to dynamically optimize emergency response plans and improve overall safety during critical events.  
The Institute of Information Science and Technologies (ISTI-CNR) plays a crucial role in developing the project's AI infrastructure through two of its research labs:  

• SI Lab: This research group develops privacy-preserving computer vision systems for pervasive indoor and outdoor monitoring. Utilizing an "AI on the Edge" approach, the algorithms detect people flows and behavioral anomalies (such as unusual crowds or blocked escape routes). To ensure full compliance with the GDPR and the AI Act, the system shares only coded mathematical information (deep features) without ever transmitting raw images.  
• KDD Lab: This research group focuses on reliable spatio-temporal Foundation Models for advanced mobility analysis. The lab generates synthetic urban mobility data to simulate complex emergency scenarios, such as floods. These generative models are essential for safely testing and optimizing hospital evacuation plans and city-scale traffic reorganization.