New Training: Functional Safety and AI Systems
SecuRESafe (SRES) is now offering a one-day training course focused on applying functional safety principles to AI-based systems operating in safety-critical environments. The course bridges established safety frameworks such as IEC 61508 and ISO 26262 with emerging AI safety guidance including ISO/PAS 8800, ISO/IEC TR 5469, and the upcoming ISO/IEC TS 22440. Participants will explore how machine learning challenges traditional deterministic safety assumptions and how modern safety engineering practices are evolving to address AI-driven system behavior.
The training examines the AI safety lifecycle, AI software classification approaches, and practical mitigation strategies such as redundancy, monitoring, supervision, and fallback architectures. Participants will also explore approaches for hazard analysis, residual risk evaluation, and validation of non-deterministic systems through statistical and system-level safety assessments for Physical AI, robotics, and autonomous systems.
Why Attend?
- Understand how functional safety principles apply to AI-based systems operating in safety-critical environments
- Explore how machine learning challenges traditional deterministic safety assumptions
- Learn how emerging guidance including ISO/PAS 8800, ISO/IEC TR 5469, and ISO/IEC TS 22440 extends established safety frameworks
- Identify AI-specific faults, hazards, and failure mechanisms across the AI safety lifecycle
- Apply verification, validation, and risk reduction concepts to non-deterministic AI systems
- Examine practical architectural mitigation strategies including monitoring, supervision, redundancy, and fallback mechanisms
- Develop a system-level perspective for evaluating AI safety in Physical AI, robotics, and autonomous systems
Key Topics Include:
- ISO/PAS 8800, ISO/IEC TR 5469, and ISO/IEC TS 22440 terminology, structure, and application context
- Application Usage Levels (AUL A–D) and Software Technology Classes (SWTC I–III)
- The AI safety lifecycle and extensions to traditional functional safety development models
- Integration of AI-related faults into hazard analysis and risk assessment activities
- Architectural mitigation strategies including monitoring, supervision, redundancy, and fallback mechanisms
- Data quality, Operational Design Domain (ODD) coverage, bias, and data drift considerations
- Statistical performance metrics, residual risk evaluation, and validation of non-deterministic systems
- Qualification and assurance considerations for AI-based development tools in safety-critical environments
Who This Training Is For
This course is intended for professionals involved in the development, integration, verification, or oversight of AI-based systems operating in safety-critical environments, including those working with:
- Functional safety activities where AI or machine learning is part of the system architecture
- AI-enabled systems for robotics, Physical AI, autonomous systems, industrial automation, or automotive applications
- Development, verification, or validation of non-deterministic software components
- Hazard analysis, risk assessment, or mitigation of AI-specific faults and failure mechanisms
- Integration of AI functionality within established functional safety frameworks such as IEC 61508 and ISO 26262
- AI-based development tools and AI-assisted engineering workflows used within safety-critical product development environments
Expert Instructors
SRES instructors deliver this training based on practical experience applying functional safety principles to AI-based systems across automotive, robotics, autonomous systems, and other safety-critical applications.
Bring This Training to Your Team
This training is currently available on request as a private, one-day course, delivered virtually or in person.
