Project 3: Design and verification of safety-critical algorithms for uninhabited aerial vehicles
Adviser(s): Natasha Neogi (Assistant Professor, Aerospace Engineering)
Project description: Engineering a safe and reliable network of Unmanned Aerial Vehicles (UAV) poses several system-related problems, not dealt with by techniques traditional to the approaches used in reliability engineering. Not only must each component be safe and reliable; it must contribute to the overall emergent properties of the system, such as safety, reliability, availability. Each subsystem must be designed and composed in order to maintain or increase the overall quality of the system. The integration of the subsystems also needs to be carried out at a local (component-based) as well as global (quality-based) scale. Furthermore, the interaction of human users must be rigorously defined in order to limit the effect of environmental and user input on the safety and performance of the system. These ideas are examined in the context of an unmanned aerial vehicle test-bed.
Student background and expected research activities:The student will be involved in the development, implementation and testing of fault detection and diagnosis algorithms for online model generation onboard the PC104 computational platform. This system will provide a redundant and fault tolerant backup to the Cloudcap Piccolo Plus autopilot. A set of closed loop controls use attitude and position information to maintain stability and navigate waypoints, and will be emulated onboard the PC104 to perform a watchdog function, and enable graceful degradation and recovery of the flight platform, in the case of a failure of the autopilot.
Students who possess a background in Hardware Design, Digital Signal Processing and Computer Programming are encouraged to apply for this position.
