Path Planning for the Autonomous Inspection of Voxelized Structures

We present an infrastructure and model inspection planning approach for known structures in unknown environments. The utilized voxel-based model representations increase the input model flexibility and allow to include environment maps of the surroundings in the planning problem. By simultaneously optimizing both coverage and path length using a GTSP representation, we are able to outperform other state-of-the-art inspection planning routines in both runtime and path length while achieving equivalent coverage.