A significant challenge for the Federal Aviation Administration (FAA) involves integrating UAS into the national air space (NAS). UAS must have self-awareness (which includes on-board system health management capabilities), awareness of environment (critical for obstacle avoidance as well as other path planning goals), and formally-specified behavioral limitations to ensure safety of people and property in the air and on the ground. Even after thorough pre-flight preparation, unexpected events may occur during UAS operation. Researchers are being challenged to develop new onboard software and hardware that detect and diagnose the faults, and enable appropriate mitigation actions. Dr. Rozier’s group has examined the System Health Management (SHM) of UAS through combining the capabilities of temporal logic runtime monitors, model-based analysis, and probabilistic reasoning. Effective SHM is a critical component of achieving safe, long-duration autonomous flight, particularly in the NAS.
LAUNCH students will learn about the SHM framework and will be responsible for creating formal specifications for onboard sensors and SHM software. Furthermore, they will build interfaces that capture real-time sensor data and test the interface using flight test data. The students will assist in pushing the frontiers in embedding SHM systems onboard for real-time decision making, thereby better enabling increased autonomy levels. They will assist in the UAS functional pattern database for automated formal specification, which is an ambitious project to utilize algorithms from artificial intelligence to generate SHM configurations for future UAS constructed with COTS components.
Project mentor: Dr. Kristin Rozier