UNMANNED AERIAL VEHICLES (UAVS) HAVE MANY APPLICATIONS SUCH AS PRODUCT DELIVERY PRECISION AGRICULTURE WILDLIFE SURVEYS SURVEYING ENVIRONMENTAL MONITORING TRAFFIC CONTROL AND INSPECTION OF PIPELINES WIND TURBINES AND SOLAR PANELS. IN-TIME PROGNOSTICS AND HEALTH MONITORING ARE CRUCIAL TO THE SAFETY OF UAVS DURING FLIGHTS. MITIGATE SAFETY RISKS UNDER HAZARDOUS CONSTRAINTS REQUIRES AN EFFECTIVE BATTERY HEALTH MANAGEMENT (BHM) SYSTEM THAT CAN PROVIDE AN EARLY ALARM THAT WARNS SYSTEM OPERATORS BEFORE THE STATE-OF CHARGE (SOC) OF A UAV BATTERY FALLS BELOW A CERTAIN THRESHOLD (E.G. 30%). IT IS IMPERATIVE TO ACCURATELY PREDICT THE END-OF-DISCHARGE (EOD) OF A UAV BATTERY THAT HAS NOT BEEN CHARACTERIZED IN A LABORATORY SETTING IN ORDER TO SUPPORT OPERATIONAL DECISION-MAKING OR PROVIDE A SUFFICIENT ENERGY BUFFER FOR LANDING MANEUVERS. THE GOAL OF THIS PROJECT IS TO IMPROVE THE SAFETY OF UAVS THROUGH DATA DRIVEN PREDICTIVE ANALYTICS. IT WILL DEVELOP A NOVEL TRANSFER LEARNING APPROACH TO PREDICTING THE EOD OF LI-ION BATTERIES OF UAVS UNDER VARYING FLIGHT PLANS AND PAYLOADS WITH BATTERY CONDITION MONITORING DATA INCLUDING CURRENT VOLTAGE AND TEMPERATURE. WE WILL VALIDATE THE PROPOSED TRANSFER LEARNING-BASED PREDICTIVE MODELING APPROACH USING THE EXPERIMENTAL DATA COLLECTED FROM A FIXED-WING UAV. THIS PROJECT IS ALIGNED WITH ONE OF THE THRUST AREAS IN THE NASA S STRATEGIC IMPLEMENTATION PLAN: IN-TIME SYSTEM-WIDE SAFETY ASSURANCE. BECAUSE IN-TIME PREDICTION OF THE EOD OF A UAV BATTERY ENABLES EFFECTIVE FLIGHT PLANNING AND IN-TIME OPTIMIZATION OF FLIGHT PLANS THE PROPOSED TECHNIQUE CAN BE EXTENDED TO URBAN AIR MOBILITY (UAM).