From a minimally-invasive medical robot automatically inspecting the surface of a diseased organ to an autonomous quadcopter inspecting the structure of a bridge, robots in a variety of scenarios must plan their motions to efficiently inspect regions of interest. These are examples of robot-inspection planning tasks which arise in diverse applications. Here, we are tasked with planning motions for a robot that enable it to efficiently inspect a region of interest using its on-board sensors. Such an inspection plan should not only inspect the region of interest but also obey the robot's kinematic constraints, avoid obstacles (e.g., parts of a bridge or sensitive anatomical structures in bridge inspection and medical applications, respectively), and minimize some metric (such as time to completion or distance traveled), all while considering real-world uncertainties (e.g., uncertainty in the robot's kinematic model or external forces). Inspection planning is extremely challenging and naively-computed inspection plans may enable inspection of only a subset of the region of interest and may be highly suboptimal.
Computing a high-quality inspection plan
is important for many applications, e.g., a medical diagnostic inspection plan should minimize procedure duration to decrease the amount of time a patient is under anesthesia, all while thoroughly inspecting the region of interest. Current approaches to computing inspection plans are either highly tailored to specific applications, provide no guarantees on the quality of solution, or come with very long computation times for non-trivial problems, rendering them impractical for real-world applications.