Uncertainty-Aware Velocity Obstacle Avoidance

The Uncertainty-aware Velocity Obstacle (UVO) algorithm seeks to guarantee the safe autonomous navigation of multiple small unmanned aircraft systems (sUAS) operating within a constrained airspace by incorporating the aircraft's estimated position and velocity uncertainty. This paper formulates a continuous 3D collision avoidance algorithm that is robust to position and velocity uncertainties. A scenario is presented to illustrate the utility of this method for an sUAS encountering other cooperating vehicles. In this scenario, the vehicles query all other vehicles' positions and velocities based on a ground radar system. Then they estimate uncertainty bounds based on an extended Kalman filter (EKF) and those bounds are used to create a collision cone that identifies unsafe areas for the vehicle. This paper provides numerical simulations and hardware experiments to validate the vehicles' ability to avoid collisions when operating with uncertainty in other vehicle's positions and velocities.