Autonomous Flying Fire Blanket: New Adaptive And Learning Architectures For Multi-UAV Cooperative FormationWith Firefighting Applications

“Autonomous Flying Fire Blanket”: New Adaptive And Learning Architectures For Multi-UAV Cooperative Formation With Fire?ghting Applications Overview: Fires cause signi?cant damage to life and the economy in the United States. Unmanned aerial vehicles (UAVs) have been increasingly used in the ?re?ghting tasks. However, current ?re?ghting UAVs suffer from several limitations and constraints, including complex designs of the platforms, limited ?ow rate and tank capacity for the suppressant ?uid, long re?ll time, collateral damage caused during ?re?ghting, and high costs. A more resilient, effective, and economically viable system will be important for the increasingly dif?cult ?re?ghting tasks. We propose to develop a new “autonomous ?ying ?re blanket”, using a team of UAVs to collaboratively tether a ?re blanket. This novel ?re?ghting system puts a high demand on the safe, precise, and resilient operations of the UAV team. Speci?cally, major technical challenges include meeting multiple formation constraint requirements on safety and performance that are time and path dependent, adaptation and learning under varying operation conditions over both the time and iteration axis, and resilient designs in the face of potential malfunctioning at the ?re scene. Current formation literature often narrowly de?ne safety as mere collision and obstacle avoidance, with little emphasis on maximum allowable inter-agent distance and attitude constraints. Both the time- and path-varying nature of the constraint requirements is often overlooked. Furthermore, existing works on adaptive or learning control frameworks only focus on learning over either the time or iteration axis. A uni?ed structure to learn and adapt over both the time and iteration domain has not been addressed in the literature. Moreover, existing cooperative formation algorithms often ignore any malfunctioning of the networked agents. To address these challenges, we will develop new adaptive and learning cooperative formation architectures based on universal barrier functions, adaptive learning structures, and resilient control framework designs, while taking external payloads, system nonlinearities, uncertainties, and external disturbances into considerations. The designed system will be indoor tested using our UAV ?eet and ?ight arena, and tested outdoors at Fort Campbell (see Letters of Collaboration). Intellectual Merit: The research outcome will signi?cantly advance current state-of-the-art in UAV cooperative control and formation architecture design that are capable of meeting multiple formation constraint requirements, adaptation and learning, and mitigating sensor and actuator malfunctioning. The proposed research aligns with the NSF Strategic Goal 1 to expand knowledge in science, engineering, and learning and the NSF Strategic Goal 2 to advance the capability of the Nation to meet current and future challenges. The Principal Investigator (PI) is uniquely quali?ed for this research effort by virtue of the expertise in adaptive and learning control, nonlinear constrained system control, and resilient control designs. Broader Impact: The PI will collaborate with Ohio Drone LLC for the software and hardware designs of the UAV team (see Letters of Collaboration). Moreover, we will work with our city ?re department, namely the Lexington Fire Department, as well as the Kentucky Division of Forestry, on the designs and veri?cations of the system for ?ghting ?res of different types, and on promoting the system for real-world ?re?ghting (see Letters of Collaboration). The proposed formation architectures can also be applied to other cooperative operations like transportation tasks. With collaboration from the Department of STEM Education in our College of Education, we will also integrate our UAV research with local middle and high school outreach (see Letters of Collaboration). We will also integrate the proposed research activity with the ongoing undergraduate and graduate teaching efforts, and involve women, minority, and students from under- represented groups in hands-on experiences related to the project.