Conservative Algorithms for Automated Collision Awareness for Multiple Unmanned Aerial Systems

As the Federal Aviation Administration (FAA) prepares to integrate Unmanned Aerial Systems (UAS) into the National Airspace System (NAS), developing technologies that mitigate the risk associated with UAS collisions have become a top priority. Despite advances in detect and avoid technologies, the UAS operator remains the primary controller responsible for maintaining inter-vehicle separation and ensuring conflicts do not occur. This paper examines a collision awareness system which increases the operator's situational awareness by spatially and temporally predicting conflicts between the UAS and entities such as other aviation traffic or restricted airspaces. By modeling entities as 3D point masses, the system can be implemented for various, dissimilar UASs. Furthermore, the system supports aircraft engaged in different flight modes such as free flight, following a flight path, and orbit/loiter behavior. Mixed Gaussian distributions model each entity's future position, where the mean is determined by 3D kinematic motion and the covariance is determined by a continuous time error propagation model. Convolving these mixed distribution with another entity or airspace yields mathematically conservative future conflict estimates. Scenarios are presented to demonstrate the algorithm's capabilities.