- SELFY toolbox solution aims at detecting over 100% security breaches.
- The first validations took place in the urban area of the ADAS/CAV test track at Applus+ IDIADA, in Catalonia, reproducing real metropolitan situations.
- Vienna hosted a live traffic demonstration in the city centre using camera sensors and roadside units distributed throughout the city.
- SELFY offers automotive manufacturers, traffic managers, fleet operators and drivers a comprehensive solution to detect, respond to and mitigate cyberattacks.
Barcelona, 26 May, 2025. The European consortium SELFY, coordinated by the technology centre Eurecat, has demonstrated the effectiveness of new self-assessment and self-protection tools designed to increase the resilience and cybersecurity of autonomous and connected vehicles in smart cities, through pilot tests held in the region of Catalonia and the city of Vienna. The SELFY solution identifies over 95% of vulnerable vehicles and more than 90% of the security breaches.
After three years of research and development, “SELFY offers automotive manufacturers, traffic managers, fleet operators and drivers a comprehensive solution to detect, respond to and mitigate cyberattacks, while fully preserving the privacy and integrity of autonomous mobility systems,” states Fanny Breuil, project coordinator and European Programme manager at Eurecat.
With this solution, “it is possible to identify vulnerable vehicles, reduce the risk of security breaches, and enable the vehicles themselves to protect and self-recover, ensuring trust and secure data exchange,” adds SELFY’s technical coordinator, Víctor Jiménez, researcher in the IT&OT Security Unit at Eurecat.
The first validation took place in the urban area of the ADAS/CAV test track at Applus+ IDIADA, in Catalonia, Spain, with a layout that replicates intersections, roundabouts, pedestrian crossings and driving lanes to faithfully reproduce real metropolitan traffic conditions.
The tests recreated scenarios such as detecting a vulnerable road user while a hacked vehicle was transmitting misleading information, identifying sensor failures through infrastructure data fusion, filtering unreliable cooperative messages, and safely aborting an overtaking manoeuvre during an active cyberattack. An anonymisation algorithm was also verified, protecting sensitive data in real time, such as pedestrian faces and vehicle licence plates.
“At Applus+ IDIADA facilities we recreated manoeuvres close to everyday traffic using dummies, real vehicles and targeted attacks. The results confirm that SELFY tools improve perception and response to security incidents without compromising privacy,” states Manel Rodríguez, Expert Engineer in cybersecurity in Applus+ IDIADA.
In parallel, in Vienna, Austria, a demonstration in the city centre was held under live-traffic conditions with camera sensors and roadside units installed across the city. After mounting a high-resolution camera and a roadside unit on the infrastructure, the SELFY system monitored the coherence between images and CAM messages and accurately detected artificially induced mismatches, validating the infrastructure’s capability to identify misaligned or tampered sensors.
“The urban pilot confirms that the same tools that work in closed tracks are equally effective in live traffic, a key step for their adoption at a European scale,” states Gernot Lenz, Coordinator Traffic Management Systems at the City of Vienna.
Intelligent systems for autonomous and connected driving
The SELFY project has developed three macro-solutions focusing on situational awareness and collaborative perception, a cooperative resilience and recovery system, and a trust and data management system.
The situational awareness and collaborative perception system fuses vehicle perception with infrastructure information to build a unified view of the environment. Its tools continuously analyse the coherence of sensors and cooperative messages to detect discrepancies and assess the reliability of each source before relying on it for driving decisions.
Meanwhile, the tools for cooperative resilience and recovery are designed to protect Cooperative, Connected and Automated Mobility (CCAM) environments against cyberattacks and security breaches. Managed by a Vehicle Security Operations Centre (VSOC), the system enhances resilience, robustness and system trust levels, and provides a secure degraded mode for vehicles when necessary.
The tools addressing trust and secure data management ensure information protection and privacy. They incorporate algorithms to detect malicious behaviour in both vehicles and roadside units, apply AI-based anonymisation techniques, and use software update mechanisms with post-quantum cryptography to ensure the integrity and traceability of each new version.
The SELFY project has been funded with €6 million under the Horizon Europe programme. The consortium, led by the technology centre Eurecat, includes partners from 8 countries, including Spain (Eurecat, Tecnalia, AEVAC, Ficosa and Applus+ Idiada); France with CEA and Canon Research Centre; Germany with Technische Hochschule Ingolstadt and FEV.io; Austria with Virtual Vehicle Research and City of Viena; the Netherlands with Eindoven University of Technology; Japan with Okayama University; Australia with RMIT University, and Turkey with FEV.
