Komatsu, Rina (Sophia University), Arntzen Bechina, Aurilla Aurelie (University of South -Eastern of Norway), Güldal, Serkan (Adiyaman University), Sasmaz Guldal, Merivan (Adiyaman University)

Machine Learning Attempt to Conflict Detection for UAV with System Failure in U-Space: Recurrent Neural Network, RNN

Scheduled for presentation during the Regular Session "U-Space Urban Air Mobility" (WeA2), Wednesday, June 22, 2022, 11:30−11:50, Bokar

2022 International Conference on Unmanned Aircraft Systems (ICUAS), June 21-24, 2022, Dubrovnik, Croatia

This information is tentative and subject to change. Compiled on April 20, 2024

Abstract

U-Space services will be used by millions of Unmanned Aerial Vehicles (UAVs) also called drones in the near future. To respond to growing demand, urban airspace needs more optimized utilization to prevent delay and collision. In this study, we have implemented a Machine Learning (ML) method to predict possible routes preventing collisions with a UAV which experiences system failure in the U-Space. Our ML implementation, by Recurrent Neural Network, provides information about the possible route to prevent unwanted events. Thus, Air Traffic Managements can reroute the necessary UAV in the U-Space environment.