Design and implementation of robotic gripper for food grasping

In this dissertation, the author designs and implements a gripper with passive adaptive fingers and flexible wrist for food grasping. The finger consists of thin half-open hollow frame and flexible string woven on the frame, which provides adaptability for different types of single food objects and the possibility of grasping closely arranged food objects, while the flexible wrist enables the gripper to grasp from any direction rather than top-down vertical grasping, which can be a solution to the grasping in random-bin picking (RBP) scenario. The open loop control of the gripper and the integration of the gripper with a Delta robot are also realized, which implements simple automation for food pick-and-place task. Certain tests are carried out, showing 10/10 grasping success rate for the food samples of peanut cake, muffin, puff, tomato, and corn segment, while 8/10 success rate for walnut. The result also shows that the gripper has the potential for grasping closely arranged food objects in rectangle layout but performs poor for hexagon layout. The flexible wrist helps the gripper to achieve 10/10 success rate for grasping an awkward-positioned object from a preset non-vertical direction, but it is found that the accuracy of joint angle control of the gripper is poor. Keywords: gripper, food grasping, adaptable finger, flexible wristMaster of Science (Smart Manufacturing