Lockton and Nexar Introduce Human Benchmark for Autonomous Vehicle Safety

Lockton and Nexar have introduced a human benchmark framework for evaluating autonomous vehicle safety against real-world human driving behavior, announced in a press release. The framework aims to serve insurers, regulators, and autonomous vehicle developers by providing an independent reference for comparing vehicle performance to human driving standards.

The framework combines Lockton's expertise in risk management and insurance with Nexar's real-world driving intelligence. It is built on Nexar's BADAS 2.0 collision anticipation model, which was trained on 10 billion miles of real driving data and 60 million safety-critical events. Nexar's network of 350,000 cameras covering 94 percent of U.S. roads provides the data foundation for establishing the human baseline.

Two initial components support the framework. The Nexar Risk Index evaluates environmental risk based on geography, conditions, and driving context before comparing system performance. Nexar Apex allows autonomous vehicle companies to test their systems against curated real-world edge cases, assessing performance relative to the human baseline.

According to the companies, the framework is designed as an open, evolving reference rather than a proprietary rating, offering a consistent method for insurers to assess risk, regulators to validate safety claims, and developers to demonstrate system performance.