Sorting parts by random grasping

As a low-cost alternative to machine vision, the authors consider how a modified parallel-jaw gripper can be used to classify parts according to shape by grasping and measuring the diameter: the distance between the jaws. Since more than one part may give rise to the same diameter and the sensor may be corrupted by noise due to surface compliance and backlash, the authors show how the most probable part can be estimated using a sequence of random grasps with a Bayesian decision procedure. This procedure allows the authors to define a statistical measure of the "similarity" of a set of parts. Laboratory experiments confirm that the random strategy is effective for sorting parts.<
>