Overview
This half-day training is an introductory course on Artificial Intelligence (AI) safety for road vehicles safety by using the state-of-the-art ISO/PAS 8800:2024 standard. It is ideal for those wanting to learn more about machine learning (ML) and AI safety and for executive management wanting a better understanding of the new and evolving standards for their organization to develop towards compliance.
This training is provided by automotive autonomy and AI safety experts with vast real-world experience working on prototype and production-ready programs. These include autonomous vehicle and robotics technology developers, global automotive OEMs, suppliers, semiconductor manufacturers and software vendors to successfully adapt their product development processes and release responsibly safe embodied AI products into the market.
Details
The training provides industry-specific guidance on the use of AI systems in safety-related functions, independent on AI methods or specific vehicle functions. This includes the usage of AI for the functionality itself or the use of AI as a safety mechanism. The objective of the training is to provide an introductory view into the critical terminology, techniques and methods for the safe development of AI systems according to the ISO/PAS 8800:2024 standard, and enable interactive discussions on the following topics:
- Introduction and Scope of ISO/PAS 8800:2024
- Relationship with ISO 26262:2018 and ISO 21448:2022 standards
- AI safety lifecycle and comparisons to “classical” development
- Structuring and evaluating assurance arguments for AI systems
- AI system development
- Derivation of AI safety requirements
- Selection of AI technologies and architectural measures
- Data-related considerations
- Verification and validation of AI systems
- Safety analysis of AI systems
- Operational measures to maintain safety of AI systems
- Confidence in use of AI development frameworks and software tools
- Discussion and Live Q&A
Objectives
The primary goal is to apprise managers and responsible engineers of the scope and implications of ISO/PAS 8800:2024 lifecycle, methods and work products for their company so they develop AI/ML systems responsibly. Particularly, the session is intended to equip the participants with a practical interpretation of the open-ended nature of the standard’s objectives and how it interplays with existing functional safety (ISO 26262:2018) and safety of the intended functionality (ISO 21448:2022) activities.
