Asymptotically optimal inspection planning via efficient near-optimal search on sampled roadmaps

Inspection planning, the task of planning motions for a robot that enable it to inspect set points interest, has applications in domains such as industrial, field, and medical robotics. can be computationally challenging, search space over motion plans grows exponentially with number interest inspect. We propose novel method, Incremental Random Inspection-roadmap Search (IRIS), computes inspection whose length successfully inspected asymptotically converge those an optimal plan. IRIS incrementally densifies motion-planning roadmap using sampling-based algorithm performs efficient near-optimal graph resulting is generated. prove under very general assumptions about environment. demonstrate IRIS’s efficacy on simulated planar five DOF manipulator, bridge Unmanned Aerial Vehicle (UAV), endoscopic continuum parallel surgical cluttered human anatomy. In all these systems higher-quality orders magnitudes faster than prior state-of-the-art method.