Safely Entering the Deep: A Review of Verification and Validation for Machine Learning and a Challenge Elicitation in the Automotive Industry
Addresses safety challenges for automotive industry adopting deep learning, but is incremental as it reviews existing methods and reports expert opinions without new technical solutions.
This paper reviews verification and validation methods for safety-critical machine learning systems in automotive applications and reports findings from workshops with automotive experts, concluding that current safety standards like ISO 26262 are incompatible with deep neural networks and recommending knowledge transfer from aerospace and system-based safety approaches.
Deep Neural Networks (DNN) will emerge as a cornerstone in automotive software engineering. However, developing systems with DNNs introduces novel challenges for safety assessments. This paper reviews the state-of-the-art in verification and validation of safety-critical systems that rely on machine learning. Furthermore, we report from a workshop series on DNNs for perception with automotive experts in Sweden, confirming that ISO 26262 largely contravenes the nature of DNNs. We recommend aerospace-to-automotive knowledge transfer and systems-based safety approaches, e.g., safety cage architectures and simulated system test cases.