Dynamic Path Planning for Avoiding Non-Cooperative Threats
Anand Pradhan Shrestha, Sabyasachi Mondal, Antonios Tsourdos · 2025
With the advent of urban air mobility (UAM) and the growth of unmanned aerial vehicles (UAVs), the skies will become increasingly crowded as we try to integrate these new air platforms in the current airspace. Therefore, robust conflict resolution and collision avoidance strategies are required to make the airspace safe for all the UAVs including the delivery drones, air taxis and medical supply drones. Adding to this challenge, there might also be non-cooperative air platforms in the airspace, which don't share any information about their position, velocity, or intent. Considering the problem of accidental interception of UAVs with non-cooperative air platforms, this work proposes a dynamic, local, and real-time path planning strategy where probabilistic methods are used for future position estimation of intruder UAVs and cubic splines to generate collision-free trajectories. Additionally, the conditions from inverse navigation has been used to determine weather the conflict resolution manoeuvre is necessary. The candidate splines in this method allow the UAVs to select an optimal path considering the smoothness and the vehicle constraints. The probabilistic method uses the Monte Carlo method to predict the probable positions of the intruder, which can be used to predict future collisions. This approach fosters safe and efficient airspace utilisation, paving the way for the successful realisation of UAM.