Abstract:
Advanced Air Mobility (AAM) is an innovative concept that aims to revolutionize air transportation through electric and unmanned aircraft, enabling applications such as urban air taxis and medical transport. However, one of the key challenges to its widespread adoption is ensuring safety, particularly in collision avoidance. This study focuses on the development of a perception and guidance system for avoiding collisions with non-cooperative targets, which do not share their position or trajectory. To achieve this, a Frequency-Modulated Continuous Wave (FMCW) radar and an InfraRed(IR) camera are used. Compared to traditional pulsed or panel radars, FMCW radars offer higher resolution, better detection of small and slow-moving objects, and improved performance in cluttered environments. The IR camera enhances situational awareness by providing visual confirmation and additional tracking capability, making this sensor fusion approach particularly suitable for AAM applications. Our collision avoidance system follows ACAS Xu standards, which provide autonomous conflict detection and resolution for unmanned aerial vehicles. The maneuver selection process is based on precomputed lookup tables generated through a Markov Decision Process (MDP), optimizing responses based on risk and energy consumption. The entire system is tested in a simulation environment using Ansys AVxcelerate, a physics-based simulator capable of generating realistic sensor data. This approach allows for comprehensive testing of detection, tracking, and maneuver execution in a highly realistic scenario, ensuring the effectiveness of the proposed solution before real-world deployment.