Safety framework for off-highway autonomous vehicle deployed piloted
Oxbotica and TRL have piloted the use of the first Code of Practice for the safe and cost-effective deployment of autonomous vehicles in unstructured off-highway environments, with a live trial conducted in a quarry using a number of off-road vehicles.
The Innovate UK-funded consortium has developed and demonstrated capabilities to adapt and retrofit autonomy, using robust low-cost sensors, to any vehicle, as well as drafting a Code of Practice that identifies the key elements for safe and efficient deployment of autonomous vehicles in off-road industries.
Off-road environments, which can include mines, quarries, farms, refineries, warehouses, ports, and airports, feature more varied hazards and less structured scenarios than on-road settings with no universal highway rules, such as speed limits or junction etiquette. A Code of Practice helps standardise across industries and allows learnings from each domain to be shared.
Off-road vehicles also have to interact with a wide variety of unpredictable objects in their environment, either because they block the vehicle’s path, such as undergrowth or tree branches, or because engaging with them is part of the vehicle’s primary function, such as harvesting or excavating.
With autonomous vehicles commonplace in many of these industries, working to deliver safety, efficiency and productivity, the Code of Practice will help organisations transition to new working practices and harness the potential benefits with no impact on safety.
To demonstrate the Code Of Practice and highlight its potential to work across a range of vehicles and industries, Oxbotica and TRL deployed a Ford Ranger and Range Rover Evoque, retrofitted with Oxbotica’s autonomy software platform, in a UK quarry in April 2021. The vehicles were fitted with a full suite of sensors, including LiDAR, RADAR, and stereo cameras.
Oxbotica’s technology has already operated in a range of environments without road markings across Europe, Asia and America. Its pioneering software seamlessly transitions between sensors to operate across multiple domains and environments. The software is capable of using sensors independently or fused in any combination, meaning vehicles can drive with or without maps, depending what is available at any given time.
