Researchers led by Pranav Nedunghat have developed a way for drones to compute minimum-time flight paths in real time while guaranteeing they avoid obstacles, according to a paper on arXiv submitted to ICRA 2027. They want drones to reach disaster areas quickly to deliver supplies or find survivors.
Earlier methods computed the fastest paths without considering obstacles, the authors write. Their method includes obstacle avoidance and counts mission time as planning time plus flight time, since both decide how fast help arrives. In an example mission, a quadrotor flies through several canyon-like obstacles to deliver aid, at close to full thrust for most of the way.



