Methodology To Develop A Discrete-event Supervisory Controller For An Autonomous Helicopter Flight

The National Research Council (NRC) of Canada is actively engaged in the development of an advanced autonomy system for the Bell 412 helicopter. This system's capabilities extend to executing complex missions, such as arctic resupply flights, where the helicopter autonomously performs tasks like takeoff with cargo, navigation while avoiding potential obstacles, and precise landing at its destination, all while minimizing the need for pilot intervention. The complexity of this autonomy system necessitates the inclusion of a high-level supervisory controller. This controller plays a critical role in monitoring mission progress, interacting with various system components, and efficiently allocating resources on board. Conventionally, supervisory controllers are embedded within monolithic programs, lacking transparent state flows and making modifications and system behavior testing a significant challenge. In our research, we present an innovative approach to develop supervisory controllers for autonomous rotorcraft. Leveraging the DEVS (Discrete Event System Specification) formalism and the Cadmium simulation engine, we effectively address the aforementioned challenges. Using the NRC's Bell-412 autonomy system as a comprehensive case study, we elucidate the entire development process for a state-based, event-driven supervisory controller for autonomous rotorcraft. This process encompasses modeling, implementation, verification, validation, testing, and deployment. It incorporates a simulation phase, in which the supervisor integrates with various components within a Digital Twin of Bell 412, and a real-time operations phase, where the supervisor becomes an integral part of the actual Bell 412 helicopter. Our method outlines the smooth transition between these phases, ensuring a seamless and efficient process.