Uncrewed Systems Technology 044 l Xer Technolgies X12 and X8 l Lidar sensors l Stan UGV l USVs insight l AUVSI Xponential 2022 l Cobra Aero A99H l Accession Class USV l Connectors I Oceanology International 2022

16 June/July 2022 | Uncrewed Systems Technology Intel has won a project to develop simulation technology for uncrewed off-road vehicle development (writes Nick Flaherty). The Robotic Autonomy in Complex Environments with Resiliency – Simulation (RACER-Sim) project has two phases over 48 months, and aims to accelerate the entire r&d process for designing such vehicles. The project is significant, as the commissioning agency, the US Defense Advanced Research Projects Agency (DARPA), has played a key role in the development of autonomous vehicles since its first challenge projects in the 1990s. In phase one of the project, the aim is to create new simulation platforms and map-generation tools that mimic complex off-road environments with the highest accuracy. This includes modelling the physics of motions, sensor modelling and building complex terrains including different types of soil and vegetation and hundreds of different types of obstacles. Creating large, complex simulation environments traditionally requires significant resources and is one of the biggest challenges in simulation workflows. Intel Labs is working with the Computer Vision Centre, in Spain, and the University of Texas at Austin on a simulation platform that will enable customisation of future maps, including the creation of new environments covering more than 100,000 square miles. It has already developed an autonomous vehicle simulation through several projects, including the open source CARLA simulator for urban environments. During phase two, Intel Labs will work with other groups in the project who are developing off-road UGVs. Carnegie Mellon University, NASA-Jet Propulsion Laboratory, and the University of Washington are using RACER Fleet Vehicles (RFVs) provided by DARPA. These have 360 º sensors with multiple Lidars, stereo camera pairs, colour and infrared imaging cameras, radar, event sensors and inertial measurement sensing. The sensors are combined with GPU processors in a rugged ‘e-box’ and collect 4 Tbytes of data per hour to support AI, machine learning-based autonomy algorithms and stack design. The RFVs were integrated by Carnegie Robotics and are housed in a Polaris RZR S4 1000 Turbo base drive-by-wire platform that has been modified for roll protection, the sensor/e-box integration, autonomous control, and a power level of 7 kW. The teams will develop new algorithms to validate the performance of the RFVs in simulation, saving significant time and resources. Phase two will also include the development of new sim2real techniques – the concept of training the robot in simulation to acquire skills and then transferring these skills to a corresponding real robotic system – enabling the training of off-road autonomous ground vehicles directly in simulation. Intel expects these new simulation tools to significantly improve the development of autonomous systems using virtual testing, reducing the risks, costs and delays associated with traditional testing and verification protocols. In the future, the simulation platform will go beyond validation to create AI models ready for implementation in the real world. Ground vehicles Off-road UGVs simulator The RACER-Sim project will simulate environments for uncrewed off-road vehicles