Sign up to save your podcasts
Or
Share Automotive Safety Standards/ISO PAS 8800
Share to email
Share to Facebook
Share to X
Share to email
Share to Facebook
Share to X
Or
Automotive Safety Standards/ISO PAS 8800
ISO/PAS 8800: Safety Standard for Artificial Intelligence in Road Vehicles
Executive Summary
ISO/PAS 8800 is a newly established international standard designed to enhance the safety of road vehicles equipped with artificial intelligence (AI) systems. Developed in response to the rapid integration of complex AI technologies in the automotive sector, the standard supplements existing safety frameworks—specifically ISO 26262 (Functional Safety) and ISO 21448 (Safety of the Intended Functionality, or SOTIF).
The standard introduces a holistic, lifecycle-based safety framework for electrical and electronic (E/E) components that utilize AI. Key takeaways include:
- Integration with Existing Standards: It bridges the gap between deterministic functional safety and the management of edge cases in AI systems.
- Data-Centric Methodology: It treats datasets as critical safety artifacts, requiring rigorous version control and traceability across five distinct categories.
- Lifecycle Management: Safety is addressed from the initial concept through development, deployment, and continuous field monitoring.
- Strategic Challenges: While providing a robust framework for risk management, the standard faces criticism regarding its ability to distinguish between system "failures" and AI "insufficiencies," as well as concerns over its practical adaptability across different industrial sectors.
View all episodes
By Veljko Massimo PlavsicISO/PAS 8800: Safety Standard for Artificial Intelligence in Road Vehicles
Executive Summary
ISO/PAS 8800 is a newly established international standard designed to enhance the safety of road vehicles equipped with artificial intelligence (AI) systems. Developed in response to the rapid integration of complex AI technologies in the automotive sector, the standard supplements existing safety frameworks—specifically ISO 26262 (Functional Safety) and ISO 21448 (Safety of the Intended Functionality, or SOTIF).
The standard introduces a holistic, lifecycle-based safety framework for electrical and electronic (E/E) components that utilize AI. Key takeaways include:
- Integration with Existing Standards: It bridges the gap between deterministic functional safety and the management of edge cases in AI systems.
- Data-Centric Methodology: It treats datasets as critical safety artifacts, requiring rigorous version control and traceability across five distinct categories.
- Lifecycle Management: Safety is addressed from the initial concept through development, deployment, and continuous field monitoring.
- Strategic Challenges: While providing a robust framework for risk management, the standard faces criticism regarding its ability to distinguish between system "failures" and AI "insufficiencies," as well as concerns over its practical adaptability across different industrial sectors.