Reducing Uncertainty Using Placement and Regrasp Planning on a Triangular Corner Fixture

This paper presented a regrasp planning method to eliminate grasp uncertainty while considering the geometric constraints of fixture. The automatically finds Stable Placement Poses (SPPs) an object on Triangular Corner Fixture (TCF), elevates from its SPPs dropping poses and Deterministic Dropping (DDPs), builds graphs by using SPP-DDP pairs their associated configurations, searches graph find motion sequences for precise assembly. Since regrasps are constrained TCF’s geometry have high precision, final regrasped via it has low can be directly used assembly position control. In experimental section, we study performance analytical learning-based methods estimating DDPs different objects quantitatively examine proposed method’s ability suppress tasks like peg-in-hole insertion sheathing tubes, aligning holes, mounting bearing housings, etc. results demonstrate robustness efficacy. Note Practitioners —In production lines, robots interact with peripheral devices improve efficiency reduce uncertainty. this work, focus particular device – (TCF) made three inclined mutually perpendicular plates. We TCF manipulation precision. plates form gravity bucket that holds dropped in stable states under gravity. real scenario, robot picks up releases above TCF. released will reach state Then, moves stabilized target pose reduced Using paper, finish procedure finding all object’s TCF, invalidating infeasible grasps, building searching sequence goal precision industrial applications, potential flexibility robotic systems high-precision tasks. research fields, may promote sensorless extrinsic manipulation, push forward studies regrasp.