Scaling the deployment of autonomous vehicles has proven challenging, largely because demonstrating their safety and acceptability to all stakeholders – including the public, regulatory authorities, and other road users – has been difficult. The widespread integration of machine learning techniques has added another layer of complexity to this challenge.

This presentation explores the multifaceted nature of autonomous driving, examining local driving conditions and the methods and processes currently in use or under deployment to validate these systems.

• Driving complexity varies significantly by location and context – there is no such thing as an “average driver” or standard driving scenario
• Validating autonomous driving functions across all real-world conditions requires diverse testing methodologies and metrics
• Verifying production-ready autonomous vehicles presents unique technical and regulatory challenges that demand rigorous processes