According to Horiba MIRA, incorporating a scenario-based testing regime into the early stages of connected and automated vehicle design and development is imperative. It is vital to improve the efficiency and cost of verifying and validating CAVs. Horiba MIRA also believes a ‘quality over quantity’ approach is essential when it comes to evaluating the number of miles completed during the development of CAV systems.
Rob Capaldi, commercial manager for connected and autonomous vehicles at Horiba MIRA, comments, “With the testing and trialling of CAVs moving into real-world deployment, it’s great to see the UK at the forefront of the market, with a large number of both physical and simulated trials.
“This can be achieved by utilizing a scenario-based testing approach, in which test scenarios are identified, categorized and managed as the key foundation to the vehicle tests. Typically, these are completed in simulation at the early stages of system development, with physical testing completed later in the development process.
“However, driving a billion miles will not necessarily tell you how safe a vehicle is. Going forward we expect the process that determines the road-worthiness of such vehicles to center less on a single test or an arbitrary number of miles traveled, and instead more on proving they can safely perform in the infinite number of scenarios that the vehicle is likely to see in the real world,” comments Capaldi.
It is often cited that billions of miles are required to demonstrate confidence in CAV systems (Rand report), which is seen as non-viable for vehicle manufacturers with multiple product types and model variants, and for technology companies with short development and product lifecycles.
This approach is already being supported by several government initiatives in which Horiba MIRA has played a key role, including the Savvy and HumanDrive projects.
As part of these programs Horiba MIRA developed unique processes to evolve vehicle and regulatory requirements into a comprehensive set of testable scenarios that will validate and verify the performance of the CAV systems. This range of test scenarios will ensure that CAVs can perform functionally and safely while driving along normally and in unexpected situations.
Horiba MIRA also revealed that while many vehicle testing programs have previously been driven by regulatory or OEM test cases, scenario-based testing is dependent on the development of a technique to automatically generate and evaluate a large number of unique scenarios as part of the development plan, and work is underway at Horiba MIRA to create tools and methods to automate this process.
Capaldi concludes, “There’s no denying that as simulation and modeling techniques become ever-more sophisticated, those firms that adopt such advances early on will not only improve the efficiency and cost of verifying and validating CAVs but also reap the benefits of better-designed vehicles in the years to come.”