SIS 2 From detection to prevention: AI-powered safety intelligence for cities

27 April 2026

Special Interest Session, Safety and resilience through intelligent systems

Special Interest Session • SIS 2 • 14:00 > 15:00 • SIS 2 From detection to prevention: AI-powered safety intelligence for cities • EMIRGAN 1 Special Interest Session
Cities face growing pressure to create safer mobility networks. This session introduces an AI-powered safety intelligence framework that helps cities move beyond traditional crash analysis toward big-data and AI-driven risk prediction. By integrating multi-source datasets—such as traffic volumes, FCD, and crash records—into unified analytical architecture, the framework enables advanced clustering, multi-factor risk assessment, and geospatial analysis. Machine learning and spatiotemporal modelling detect high-accident-density areas, identify contributing factors, and generate dynamic risk scores. These insights support the transition toward CCAM, enabling multi-layered safety approaches, VRU-focused strategies, and scenario-based planning. The session also explores proactive safety modelling, resilience-based network analytics, and multi-source data fusion, helping cities anticipate emerging risks before crashes occur and evaluate network resilience under disruption scenarios. Complementing predictive analytics, the session presents an operator-centric perspective on how modern traffic control systems can transform control rooms into resilient 24/7 decision hubs. Drawing on deployments in European motorway and urban corridors, it demonstrates the integration of incident detection, C-ITS/V2X services, and playbook-based workflows within a unified operational environment. Real-world examples—including Türkiye’s nationwide safety analytics program and the EIT Urban Mobility project with TU Berlin, Konya, and Sarajevo—show how predictive intelligence and coordinated operations enhance resilience and enable evidence-based safety decisions.

• 223 •