UK mapping agency Ordnance Survey (OS) has revealed its role in developing a world-first simulation model based on artificial intelligence (AI) that will be used for testing connected and autonomous vehicle (CAV) safety.
OS is one of 11 partners in OmniCAV, a £3.9m (US$4.9m) government-backed project to develop a world-first AI-based simulation tool that will be used to create and run different scenarios for the safe testing of CAVs. The testing certification model can then be used by accreditation bodies, insurers and manufacturers to accelerate the deployment of self-driving transport.
The project is an important step toward the development of Geospatial Digital Twins, digital operational models of the real world that will support and enable infrastructure and mobility using data and information from a number of sources, often in real time.
OmniCAV is led by Latent Logic, a young UK startup developing AI within virtual environments, and other consortium partners include: Aimsun Limited; Arcadis Consulting (UK) Limited; Arrival Limited; Eui Limited; Oxfordshire County Council; Thatcham Research; UKAEA; University of Warwick; and XPI Simulation Limited.
OS will lead the capture, processing and serving of high-resolution geospatial mapping data that will be key for accurately testing CAVs in a digital environment. This will include 3D geometry and information about the roadside assets and their characteristics, so that data standards and requirements can be developed for the future real-world deployment and operation of CAVs.
The simulation environment will feature a 20 mile-long (32km) circuit of Oxfordshire roads, covering rural, urban, main roads and intersections. OS used similar techniques in projects such as the UK’s CityVerve IoT demonstrator, and in Bournemouth as a test case to produce a 5G planning and roll-out tool.
The company will also draw on the experience of its Atlas and E-CAVE initiatives that are helping the government shape a national infrastructure capable of supporting a nationwide network of CAVs.
“OmniCAV is one of a number of government-sponsored projects that supports the position of the UK as the best place in the world for autonomous vehicle testing. Building on the unique strengths of our capability and data, the aim for OmniCAV is to help accelerate the safe deployment of autonomous vehicles on UK roads by permitting testing in linked versions of the real-world environment and a digital one,” explained Miranda Sharp, head of innovation at OS.
“To achieve the CAV ambition, geospatial data is key. Not just to the creation of a realistic simulation environment to test in, but also to the eventual deployment of a new infrastructure to enable improved connectivity and mobility.”
Kirsty Lloyd-Jukes, CEO of Latent Logic, said, “OmniCAV’s vision is ‘CAVs for All’: bringing safer, smarter, self-driving mobility for urban and rural areas. But first we need to know that driverless cars really can handle our challenging road conditions, on country lanes as much as crowded city streets.
“Virtual reality ‘driving tests’ are the only way of doing this, which is why we’ve brought together these 11 leading organizations to build a world-first, AI-based simulation of real Oxfordshire roads to securely and reliably test autonomous car safety.”