A novel object slicing-based grasp planner for unknown 3D objects

Abstract

Robotic grasp planning has been one of the active areas of research in robotics but still remains a challenging problem for arbitrary objects even in completely known environments. Most previously developed algorithms had focused on precision/fingertip type of grasps, failing to solve the problem even for fully actuated hands/grippers during enveloping/adaptive/wrapping/power type of grasps, where each finger makes contact with an object at several points. Kinematic closed-form solutions are not possible for such an articulated finger, which simultaneously reaches several given goal points. This paper presents a framework for computing the best grasp for robotic hands/grippers, based on a novel object slicing method. The proposed method quickly finds contacts using an object slicing technique and uses a grasp quality measure to find the best grasp from a pool of pre-grasps. The pool of pre-grasps is generated by dividing the objects into parts and organizing them in a decomposition tree structure, where the parts are approximated by simple box primitives. To validate the proposed method, the developed grasp planner has been implemented on an industrial Motoman robot and a two-finger gripper. Further, the results have been compared with the state-of-the-art in grasp planning to evaluate the performance of the proposed grasp planner. As compared to other existing approaches, the proposed approach has several advantages to offer. It can handle objects with complex shapes and sizes. Most importantly, it works on both point clouds taken using a depth sensor and polygonal mesh models. It takes into account hand constraints and generates feasible grasps for both adaptive/enveloping and fingertip type of grasps. KeywordsGrasp planner • Spatial search trees • Grasp quality metric • Underactuated robotic hand/gripper B Sainul Islam Ansary

Semantic Trajectory Planning for Industrial Robotics

“…Nowhere is this more salient than in applications like pick-and-place operations, wherein robots are tasked with the precise location and grasp of specific objects. The study by (Ansary et al, 2021) emphasizes the significance of object recognition in empowering autonomous service robots to navigate and interact effectively with real-world environments (Ansary et al, 2021;Ren et al, 2021).…”