Provably Constant-time Planning and Replanning for Real-time Grasping Objects off a Conveyor Belt

In warehouse and manufacturing environments, manipulation platforms are\nfrequently deployed at conveyor belts to perform pick and place tasks. Because\nobjects on the conveyor belts are moving, robots have limited time to pick them\nup. This brings the requirement for fast and reliable motion planners that\ncould provide provable real-time planning guarantees, which the existing\nalgorithms do not provide. Besides the planning efficiency, the success of\nmanipulation tasks relies heavily on the accuracy of the perception system\nwhich is often noisy, especially if the target objects are perceived from a\ndistance. For fast moving conveyor belts, the robot cannot wait for a perfect\nestimate before it starts executing its motion. In order to be able to reach\nthe object in time, it must start moving early on (relying on the initial noisy\nestimates) and adjust its motion on-the-fly in response to the pose updates\nfrom perception. We propose a planning framework that meets these requirements\nby providing provable constant-time planning and replanning guarantees. To this\nend, we first introduce and formalize a new class of algorithms called\nConstant-Time Motion Planning algorithms (CTMP) that guarantee to plan in\nconstant time and within a user-defined time bound. We then present our\nplanning framework for grasping objects off a conveyor belt as an instance of\nthe CTMP class of algorithms.\n