Flight Test Analysis of UTM Conflict Detection Based on a Network Remote ID Using a Random Forest Algorithm

In an area where unmanned aerial system (UAS) traffic is high, a conflict detection one of the important components for safety UAS operations. A novel management (UTM) monitoring application was developed, including using inverted teardrop based on real-time flight data transmitted from network remote identification (Remote ID) modules. This research aimed to analyze performance UTM-monitoring test statistical and machine learning approaches. The tests were conducted several types small fixed-wing vehicles (UAVs) controlled by human pilot Taiwan cellular communication in suburban rural areas. Two scenarios that involved stationary, on-the-ground intruder flying used simulate event. Besides method calculating mean standard deviation, random forest algorithm, regressor classifier modules, parameters timing tests. result indicates processing time UTM most significant parameter warning parameter, besides relative distance height between UAVs. addition, latency higher than also position ground position. findings our study can be as reference aviation authorities other stakeholders development future systems.