Comments on GRVA-11-03

Proposal to establish an informal working group to perform a literature review of existing available standards, research, best practice and horizontal regulations relating to the use of artificial intelligence, develop a best practice guidance document for the use and operationalization of artificial intelligence such as machine learning in automotive safety systems, report back to WP.29 with a recommendation for publication of the guidance, and consider the need for further work. Manufacturers present vehicles developed with artificial intelligence features performing safety critical functions. Approval authorities and their technical services lack a frame of reference for auditing such development processes. Horizontal artificial intelligence regulations being developed regionally may not suit automotive sector needs.

Proposal to insert recommended requirements for AI implementation in regulated automotive safety systems based on AI-08-08, AI-06-11, and AI-04-09 documents. The recommended requirements address identification and classification of risk levels, derivation processes for general and AI-specific requirements, requirements regarding datasets, definition of test criteria and thresholds, iterative evaluation processes, and continuous processes throughout vehicle lifetime. Safety measures for AI systems must be implemented with severity appropriate to potential risks, with gap analysis conducted against design specifications to identify AI-specific risks not covered by existing standards.