World-first Artificial Intelligence-based simulation model for testing autonomous car safety kicks off in Oxfordshire

Residents in Oxfordshire may have recently seen these unusual vehicles driving on their roads, but what are these vehicles doing?

As part of its vision to enable a fully connected digital Britain, OS is collaborating with ten partners in a UK government-funded project called OmniCAV. The project plans to shape the way we will one day travel – in a connected and autonomous way – and the vehicles are collecting road and roadside data to help enable that vision.

The data OS is collecting will be made into a 3D model that will be entered into XPI’s simulator software to act as a virtual proving ground for running test scenarios. This high-fidelity model of real-world roads will be populated with realistic artificial intelligence (AI) based road users, trained with traffic camera data. Analysis of accident data, near misses and insurance data will be used to create an extensive open-access library of virtual reality simulator scenarios to test connected and autonomous vehicles.

The principal aim is to achieve data standards and requirements that can be developed to enable the real-world deployment and operation of Connected Autonomous Vehicles (CAVs). It is hoped the creation of a certification process that incorporates the virtual test environment and the subsequent real-world testing will accelerate the safe deployment of CAVs onto UK roads.

Simon Navin, Ordnance Survey’s Head of Innovation Programmes, said: “Testing in the simulation allows the project team to operate in safety without any consequence to the real world. The project will culminate in a CAV being put through the entire end-to-end OmniCAV testing programme, from simulator-only, to controlled environment, to on-road testing. Through representation on international standard committees, OmniCAV’s results will influence, or lead to the creation of, new international standards to ensure safe deployment and certification of CAVs.”

To accelerate the validation of CAV systems in the real-world, a vast number of scenarios can be tested virtually by the OmniCAV team. The digital environment allows the project team to learn quickly and in safety, enabling a rapid understanding of the behaviour of CAVs. Oxfordshire is ideal for this project because of its convenient mix of rural and busy urban roads and difficult intersections where incidents are most likely to happen.

If successful, OmniCAV hopes to scale this work internationally to establish the UK as a global leader in simulation-based certification of CAV.