Show more
Model-Centric Safety Assurance for AI-Driven Autonomous Vehicles
Learn how AI-driven autonomous vehicles are assured for safety through requirements, evaluation, verification, runtime monitoring, and standards-based evidence.
Micro-credential developed in partnership with
Model-Centric Safety Assurance for AI-Driven Autonomous Vehicles focuses on how safety is established and maintained for AV AI systems once they are placed in a safety-critical context. It explains how safety requirements, acceptance criteria, KPIs, stress testing, verification and validation, runtime monitoring, fallback strategies, and safety cases work together to support safe operation within the Operational Design Domain (ODD). The course emphasizes how evidence is gathered, interpreted, and used to justify that AI-enabled vehicle behaviour remains acceptably safe across development, deployment, and ongoing operation. It also connects this assurance view to standards and governance topics such as ISO 26262, ISO 21448/SOTIF, ISO/PAS 8800, organizational safety management, and public trust.
Micro-credential developed in partnership with
Model-Centric Safety Assurance for AI-Driven Autonomous Vehicles focuses on how safety is established and maintained for AV AI systems once they are placed in a safety-critical context. It explains how safety requirements, acceptance criteria, KPIs, stress testing, verification and validation, runtime monitoring, fallback strategies, and safety cases work together to support safe operation within the Operational Design Domain (ODD). The course emphasizes how evidence is gathered, interpreted, and used to justify that AI-enabled vehicle behaviour remains acceptably safe across development, deployment, and ongoing operation. It also connects this assurance view to standards and governance topics such as ISO 26262, ISO 21448/SOTIF, ISO/PAS 8800, organizational safety management, and public trust.
Model-Centric Safety Assurance for AI-Driven Autonomous Vehicles
Topics
Autonomous Vehicle, Hazard Assistant, ML Model
Intermediate
Price:
Included in subscription
Time to complete:
30 hours
Outcome
+3000 points
What you'll learn
-
Explain how the AV stack and the Operational Design Domain (ODD) shape safety expectations for autonomous vehicle operation
-
Connect hazards, HARA, safety goals, and acceptance criteria to concrete safety requirements for AV systems
-
Evaluate AI model behaviour using safety-oriented concepts such as missed detections, false alarms, latency, uncertainty, confidence, and residual risk
-
Interpret how safety KPIs and acceptance thresholds link model behaviour to system-level safety outcomes
-
Describe how stress testing, scenario-based simulation, hardware-in-the-loop, integration testing, and regression testing contribute to AV safety assurance
-
Understand how runtime monitoring, anomaly detection, drift handling, and fallback strategies support continuous assurance after deployment
-
Explain how safety cases are built and updated using test evidence, field data, incidents, and near-misses
-
Recognize how standards, governance, compliance, and public trust shape the safe deployment of AI-driven autonomous vehicles
Program outline
Developed with top post-secondary institutions and leading organizations, earn a credential you can share online by completing this course.
Industry-recognized
Downloadable certificate
Model-Centric Safety Assurance for AI-Driven Autonomous Vehicles
