Stephan Myschik

Christian Wittke, Stephan Myschik, Oliver Niggemann

We investigate conditional invertible neural networks (cINNs) as probabilistic inverse-dynamics models for multirotor control. For a planar X8 coaxial multicopter, we learn $p(u \mid s_t, c_t)$ from an incremental nonlinear dynamic inversion (INDI) teacher using rational-quadratic spline coupling and invertible linear mixing. Open-loop reproduction reaches $R^2 = 0.944$, mean CRPS 0.0915, and log-probability-error correlation $ρ= -0.60$. Over 15 closed-loop scenarios, position RMSE matches INDI (9.7 vs. 9.5 m), with 47 percent tracking acceptably; failures separate into attitude divergence under aggressive steps and phase lag under high-frequency references, isolating command bandwidth and data coverage as dominant failure mechanisms.

Konstantin Dmitriev, Shanza Ali Zafar, Kevin Schmiechen et al.

The emergence of a global market for urban air mobility and unmanned aerial systems has attracted many startups across the world. These organizations have little training or experience in the traditional processes used in civil aviation for the development of software and electronic hardware. They are also constrained in the resources they can allocate for dedicated teams of professionals to follow these standardized processes. To fill this gap, this paper presents a custom workflow based on a subset of objectives derived from the foundational standards for safety critical software DO-178C/DO-331. The selection of objectives from the standards is based on the importance, degree of automation, and reusability of specific objectives. This custom workflow is intended to establish a lean and highly automated development life cycle resulting in higher quality software with better maintainability characteristics for research and prototype aircraft. It can also be proposed as means of compliance for software of certain applications such as unmanned aircraft systems, urban air mobility and general aviation. By producing the essential set of development and verification artifacts, the custom workflow also provides a scalable basis for potential future certification in compliance with DO-178C/DO-331. The custom workflow is demonstrated in a case study of an Autopilot Manual Disconnection System.