Minimum time trajectory generation for surveying using UAVs

In this paper, we present a motion planning strategy for UAVs that generates a time-optimal trajectory to survey a given target area. There are several situations where completing an aerial survey is time sensitive, such as gaining situational awareness for first responders, surveying hazardous environments, etc. One of the challenges in such cases is to plan a time-optimal trajectory for the drone. To this end, we present an autonomous aerial survey framework that minimizes the time taken to completely explore a target area or volume using drones. In this work, (i) we present an approach, where for a known flight survey pattern, the planner can generate time-optimal flight paths in 3-D (ii) we frame the planning problem as a discrete non-linear program, and reduce the time taken to compute its solution by using an SOCP relaxation (iii) The given path is then executed using a simple trajectory tracking controller on a quadrotor to demonstrate its capability on hardware.