iDriving focuses on enhancing road safety through intelligent, data-driven solutions for urban and secondary roads across Europe - combining AI, digital twins and real-time monitoring with deep collaboration with municipalities and policy-makers.
Urban and secondary roads bear the brunt of growing complexity: more vulnerable users, ageing infrastructure, sudden weather events, and an evolving mix of vehicles. iDriving brings beyond state-of-the-art research to bear on this challenge - detecting hazards, optimising maintenance, and enabling efficient vehicle-to-infrastructure communication.
The project's objectives include creating a Safety Criteria Catalogue, improving traffic flow, reducing accidents, and promoting sustainable road usage. iDriving emphasises collaboration with municipalities and policy-makers, fostering solutions aligned with end-user needs and European safety standards.
Excerpts from D1.5, the first periodic progress report (submitted 23 December 2025).
Enhanced SCC built on EU 2021-2030 safety policy KPIs covering speeding, seatbelt and helmet use, alcohol, mobile distraction, vehicle safety, infrastructure and post-crash care - aligned with ISAD.
YOLOv11 models trained on 4,000-image datasets across 15+ classes; ~85-87% mAP, 11-12 FPS on NVIDIA Jetson edge; DeepSort & ByteTrack tracking; OCR-based licence-plate recognition.
SUMO + CARLA co-simulation for surrogate safety analysis; Nevers SUMO model calibrated with real traffic data; 3D Gaussian Splatting for scene reconstruction.
HTTP-based interoperability validated across components; Kafka backbone under integration; Tekniker Dataspace Connector developed; usage-control policy framework defined.
Five-day on-site trials in February 2026 confirmed real-time hazard & violation detection. Nevers (PUC 1.2) rehearsal completed 24 Sept 2025.
Active in the EU Road Safety Cluster (8 sister projects), Integrated CCAM Technology Cluster, plus joint workshops with sister Horizon Europe projects CAMBER & EvoRoads.
Develop a Safety Criteria Catalogue for urban and secondary roads, setting benchmarks for safety standards and KPIs to guide hazard identification, mitigation, and user alerts.
Implement innovative Vehicle-to-Infrastructure (V2I) communication protocols to improve real-time data exchange between infrastructure elements and diverse road users.
Use AI-powered edge computing to detect and anticipate road hazards - defects, traffic incidents - with minimal latency, enabling swift action to enhance road safety.
Establish a Digital Twin system to simulate road conditions and maintenance needs, allowing proactive, cost-effective interventions that improve infrastructure resilience.
Conduct real-world trials and user training sessions to validate the iDriving system's performance and ensure alignment with user needs and safety standards.
Promote Open Science by sharing data and publications through open-access platforms - broadening access to project results and fostering research-community collaboration.
Project Coordinator & Technical Manager
Information Technologies Institute · Centre of Research & Technology - Hellas (CERTH)
Project Scientific Manager
Technical University of Crete (TUC)