A Regulatory and Technological Framework for the Safe and Responsible Deployment of Advanced Driver Assistance Systems
Dušan Zalar, Member of ITS Slovenia, DCL and DUJPP Advisory Board

The deployment of advanced driver assistance systems with an extended operational design domain (Extended ODD), the so-called L2++ sub-level, on public road networks represents a technological leap that, due to discrepancies with existing legislative frameworks, generates substantial safety and legal risks. Although these systems operationally perform the entire dynamic driving task (DDT) in complex urban environments, legal liability and the obligation to supervise remain exclusively with the driver. This creates a structural informational and legal asymmetry: manufacturers use the public road network to optimise algorithms, while individual drivers bear all operational risks during system failures. An additional research challenge arises from the lack of sensor redundancy under conditions involving unadapted local infrastructure (infrastructural edge cases) and from the telemetry logging methodology, which frequently fails to classify anomalies as system errors when drivers intervene in time, thereby distorting the real-world assessment of safety risks.

To address the aforementioned challenges, this paper proposes a comprehensive technical and regulatory framework that mandates the implementation of a cloud-based DSSAD (Data Storage System for Automated Driving) back-end infrastructure as a prerequisite for vehicle type approval. The framework is based on the Data Escrow Storage principle, which, in the event of critical kinematic or system-related events, ensures cryptographically verified and standardised logging of parameters via a unified Application Programming Interface (API). Such an approach ensures evidentiary symmetry by shifting the burden of proof regarding software integrity onto manufacturers and enables forensic traceability for independent traffic accident investigations. The proposed back-end system grants infrastructure managers anonymised access to data for the remediation of hazardous road sections, whilst providing recognised scientific institutions with access to data to foster public safety, artificial intelligence development, and scientific transport research.