Project A06: Dynamic Consistency Management for Autonomous Traffic Agents

This project introduces dynamic consistency management for Automated Traffic Agents (ATAs), extending the CRC’s static design-time consistency to run-time. It proposes a methodology to build and maintain a database of formalized conflict clusters, enriched with resolution strategies and causal explanations, using sources like traffic rules, simulations, and accident data. At design time, V-SUMs ensure consistent ATA behavior across sub-models. At run-time, when environmental changes cause inconsistencies—situations where no rule- and goal-preserving action exists—the system identifies the conflict, selects an optimal resolution, and updates the conflict database dynamically. This approach enables ATAs to adapt to unforeseen scenarios while preserving safety and compliance, and provides explainable conflict resolution for stakeholders, ensuring transparency and trustworthiness.