Issue 39 Unmanned Systems Technology August/September 2021 Maritime Robotics Mariner l Simulation tools focus l MRS MR-10 and MR-20 l UAVs insight l HFE International GenPod l Exotec Skypod l Autopilots focus l Aquaai Mazu
36 Focus | Simulation tools of them give 40 million potential scenarios, but the design exploration and optimisation tools in the simulation software can identify potential failures such as a crash. The tools identify the parameters that lead to the failures, reducing the 40 million scenarios to 40,000 crashes and several dozen that are critical to optimising the sensor, central controller and braking system. The idea is to focus the generation of this synthetic data on the key use-cases, to reduce the time a car has to spend being road tested. The first simulation systems used gaming engines to develop the environments. That makes sense to an extent, as there are tools to help games developers build virtual worlds, and the engineers used those to build virtual versions of real-world highways for testing. The games engines are also built to support a lot of different elements moving independently, so the engineers August/September 2021 | Unmanned Systems Technology Researchers at the Massachusetts Institute of Technology have devised a mathematical representation called a probabilistic graphical model that can be used as the foundation for predictive digital twins that can handle fleets of vehicles. Custom digital twin implementations typically require a lot of computing resources, which is a barrier to real- world deployment. This is exacerbated by the fact that digital twins are most useful in situations where there are many similar vehicles, such as a fleet of UAVs. The probabilistic graphical model has been designed for creating digital twins for an entire fleet of aircraft. Expanding the use of digital twins to guide the flight of hundreds of UAVs in a swarm requires a different approach from the one-off, highly specific digital twins that are usually created to be part of a design or a simulation The model mathematically defines a pair of physical and digital dynamic systems, coupled together via two-way data streams as they evolve over time. In the case of a UAV, for example, the parameters of the digital twin are first calibrated with data collected from the physical version of it so that its twin is an accurate reflection from the start. As the overall state of the UAV changes over time through processes such as mechanical wear and tear and flight time logged, these changes are observed by the digital twin and used to update its own state so that it matches the physical UAV. This updated digital twin can then predict how the UAV will change in the future, using this information to optimally direct the physical asset going forward. The graphical model allows each digital twin to be based on the same underlying computational model, but each physical asset must maintain a unique ‘digital state’ that defines a unique configuration of the model. That makes it easier to create digital twins for a large collection of similar physical assets. The model has been tested with a UAV with a 4 m wingspan developed by Aurora Flight Sciences, and fitted with sensors from the Jessara Group that were used to collect strain, acceleration and other data. The UAV was the test bed for everything from calibration experiments to a simulated ‘light damage’ event. Its digital twin was able to analyse sensor data to extract damage information, predict how the structural health of the UAV would change in the future, and recommend changes in its manoeuvring to accommodate those changes. This idea of maintaining a persistent set of computational models that are constantly being updated and evolved alongside a physical asset over its entire life cycle is really the essence of digital twins. Boosting UAV simulation in a fleet of digital twins A simplified model enables a digital twin of UAVs in a swarm (Courtesy of MIT)
Made with FlippingBook
RkJQdWJsaXNoZXIy MjI2Mzk4