UVigo coordinates European project to prevent traffic accidents with AI

The Road4Safe project will create a digital twin of roads and traffic to anticipate risks, with pilots in Austria and Biscay.

Generic image of data flows and risk prediction on a road network with red hotspots.
AI

Generic image of data flows and risk prediction on a road network with red hotspots.

The University of Vigo, through the GeoTech group, is coordinating a consortium of 18 partners from 12 countries in the Road4Safe project, which aims to improve road safety by predicting and preventing accidents using artificial intelligence and big data.

The Road4Safe project will develop a "digital twin" (DT) of road traffic and infrastructure. This twin will integrate data from multiple sources, such as connected vehicles, infrastructure sensors, environmental data, and historical records, to generate real-time risk predictions and hazard maps at local and network levels.
Two pilot projects will be carried out: one in Austria, focusing on critical scenarios in tunnels and entry/exit ramps, and another in the Biscay Corridor (Spain), which will evaluate risk prediction at a network scale. The ultimate goal is to offer a comprehensive solution for proactive safety management, aligned with the EU's "Vision Zero" strategy.
Professor Pedro Arias, a researcher from UVigo's GeoTech group, highlights that the main focus is prevention, especially on secondary roads and transition sections like highway access points or tunnels, where speed differences and less standardized design increase accident severity. In 2023, nearly 25,600 people died on EU roads, 56% of them on secondary roads.
The project will address real-time traffic management (anticipating congestion, detecting erratic driving, collision warnings), infrastructure analysis (identifying risk zones, simulating maintenance measures), and connected mobility (vehicle-infrastructure communication, interface design for warnings).
The consortium comprises 18 partners from 12 countries, including universities like the University of A Coruña, Delft, and Queensland, and technology companies such as HAAS Alert and Commsignia, as well as Cemosa and Xouba Ingeniería. UVigo coordinates the project, funded by the Horizon Europe program with a budget exceeding 5 million euros, of which approximately 862,000 € are allocated to the university.
AI will be key to integrating and cleaning fragmented data, predicting risks at different scales (network risk maps and real-time predictions for specific sections), and supporting decision-making for road managers. It will also be used to transmit warnings to drivers through cooperative intelligent transport systems.
The reliability and explainability of the AI will be ensured, avoiding biases and complying with European data protection and artificial intelligence regulations. The pilots will validate the technologies in the Rustenfeld tunnel (Austria) and the Biscay Connected Corridor.
Based on information from the official source: Universidade de Vigo — DUVI (Diario da Universidade) (02/11/2026)