Developing Advanced Drone Navigation Algorithms at Continental and Elektrobit

This presentation shows practical insights into the application of Model-Based Design for the development of drone and eVTOL (electric vertical takeoff and landing) systems, with a particular focus on navigation enhancements. The joint project between Continental and Elektrobit addresses the critical need for robust navigation in unmanned aerial systems, especially in scenarios where GNSS signals are unavailable. The project leverages advanced technologies like GPS, inertial navigation systems (INS), and automotive radar sensors for accurate positioning. The extended Kalman filter algorithm is utilized for accurate data fusion. Continental and Elektrobit integrate a legacy automotive positioning algorithm into the systems model, demonstrating the flexibility for code reuse in Model-Based Design. ROS Toolbox was used for broad system simulation, including middleware, to pinpoint and mitigate early-stage development vulnerabilities. By adopting a layered software architecture and standardizing interfaces, they minimized the time needed for interface customization. This strategy has streamlined the process from simulation to code generation and execution. To validate the navigation solution, real-world data was leveraged in simulation. This ensured the reliability and robustness of their system under diverse operational conditions. MATLAB^® and Simulink^® products provided a reliable modeling and simulation framework and were instrumental throughout all the development phases.