Immense Simulations to inform autonomous fleet control

Transport Systems Catapult (TSC) spin out Immense Simulations and Improbable will collaborate to create a new solution for the coordination of autonomous vehicle fleets.

Immense Simulations will develop ‘tools for autonomous logistics operations and management’ building on Improbable’s SpatialOS distributed simulation platform. SpatialOS is the new operating system that enables developers to build massive, detailed simulations running on thousands of machines in the cloud.

The project team will build on existing research in cooperative routing, fleet operations, predictive vehicle health management and real time traffic management. By combining the latest thinking with Improbable’s SpatialOS developer platform, fleet operators will be able to optimise the operation of autonomous fleets at city-wide scale, for the first time.

TSC CEO Steve Yianni said, ‘Fleet operations and logistics planning is a well-established element of the value chain for fleets of vehicles. It is expected that autonomous vehicles will have a huge impact in this area, and that this project will help the industry prepare for this eventuality by developing solutions for operating fleets of autonomous vehicles. Transport Systems Catapult spin-out ‘Immense Simulations’ is designed to meet the needs of this project.’

The Collaborative Research and Development project team also includes Cubic Transportation Systems, a leading integrator of payment and information technology and services for transportation authorities and operators. The research work will be performed at the Cubic Innovation Centre in London and in Transport Systems Catapult Milton Keynes office.

Herman Narula, CEO of Improbable, said ‘Research in autonomous vehicles has so far centred on the technical problem of removing drivers. This ambitious project will fill a much needed gap; how we then optimise these vehicles on a huge scale to ensure they fulfil the promise of greater efficiency. Equally, little effort has been made in understanding the indirect consequences of introducing the autonomous vehicles. The ability to simulate the potential effects of these fleets will be invaluable to the industry.

‘Powerful simulation which includes thousands, or even millions, of entities simulated in real-time, has always been hindered by the inability of developers to integrate and scale existing models easily, or to run them at massive scale or in real time. Our platform, SpatialOS, solves these problems. We are excited by the ambition of this project and look forward to seeing how SpatialOS will enable a new tool for the transport industry.

The OECD has predicted that the autonomous vehicles fleet will deliver the same mobility with 70% fewer vehicles, but that they will drive further than the current fleet. It is through fleet optimisation, then, that the opportunity arises to ensure that autonomous vehicles have a positive impact on pollution and congestion, as well as creating significant savings for fleet operators.