Modular Modelling of Ground and Air Risks for Unmanned Aircraft Operations Over Congested Areas

Unmanned Aerial System (UAS) operations in Europe are possible under the so-called Open, Specific and - in the future - Certified Category. The Specific Category is inherently coupled with the Specific Operation Risk Assessment (SORA) as Acceptable Means of Compliance (AMC). We leverage the existing methodology of SORA as an AMC to propose a novel method for high risk operations over congested areas with a modular data-driven approach. Due to the amount of data and risk classes involved, this is a difficult problem, which requires the fusion of the available information in order to generate feasible solutions. Hence, we propose an approach, which employs heterogeneous geospatial data sets from dissimilar sources to derive metrics for operational risk. Aircraft and mission specific parameters, as well as regulatory requirements are modeled into each risk layer. This process allows for highly accurate and multi-dimensional models of risks associated with the intrinsic mission parameters. As a potential application, we evaluate the effect of high-dimensional risk models on the UAS path planning process of risk-minimal paths in realistic scenarios. We demonstrate the proposed method using commonly available APIs to derive 3D risk maps based on a variety of data classes. Test results show that comprehensive mission and vehicle specific risk data bases covering areas greater 100 km₂ can be generated on consumer hardware within a sub-hour timeframe.