A Review on AI Safety in Highly Automated Driving

Abstract

This systematic review examines approaches for assessing the safety of artificial intelligence systems used in highly automated driving. The authors investigate established safety methods as well as newer approaches developed specifically for machine learning and AI systems. Particular attention is given to validation, verification, testing, robustness, and methods for determining whether AI systems can operate safely in safety-critical environments.

Description

A systematic review of AI safety methods for highly automated vehicles. The paper examines traditional safety standards alongside newer AI-specific approaches to validation, verification, testing, and robustness.

Authors

Keywords

Comments (0)