The Virtual Testing System developed by the Simulator Office of OBRUM is a tool that plays an important role in development of autonomous systems for land vehicles. The initial design jobs and further actual engineering of the system has faced the designers to some challenges, such as access to a test platform, expenses necessary to adapt a vehicle to initial investigation of the system or availability of training and test data for the algorithm of artificial intelligence. It is why the idea was suggested to develop a system capable of resolving the encountered problems by offering the simulation environment of any land platform with on-board equipment including both passive and active sensors of various types and also various types of actuators or manipulators.
The key issue for OBRUM is safety of offered solutions and therefore commencement of any tests associated with autonomous driving of an actual platform is preceded with long-term and sophisticated tests with the use of the Virtual Testing System. Our system enables verification of software component deployed on any level of the system abstraction, i.e. not only the data acquisition or data preprocessing modules are tested, but also more sophisticated tasks are verified, such as location of a vehicle in 3D space, determination of the travelling path or classification of objects and understanding of the ambient world.
Key functionalities of the Virtual Testing System (WST) include:
• Simulation of physical behaviour of a wheeled or tracked platform in an advanced graphic 3D environment,
• Detailed reproduction of the surrounding space where a robot is to be operated, both open space and confined areas, with the use of such tools as photogrammetry of LiDAR measurements,
• Simulation of any sensor types, including LiDAR, IMU, GPS, RGB CCTV and stereovision cameras,
• Interoperability of simulation data and compatibility with actual autonomous systems.
The WST system can also be used to train future operators of robots or other unmanned vehicles. Owing to seamless integration of the simulator with the system of autonomous driving developed by OBRUM, future users can easily and fully understand the underlying principles of unmanned vehicles control and to learn how use control applications for equipment operators.
