Enabling Fixtureless Assemblies in Human-Robot Collaborative Workcells by Reducing Uncertainty in the Part Pose Estimate

Abstract

This paper presents an automated part localization technique for performing fixtureless assembly operations in a human-robot collaborative workcell. The main focus of this paper is to localize the part in the least possible time and ensure the part is not damaged during the contact-based probing phase due to the uncertainty in the preliminary pose estimation. Given the preliminary estimate of the part pose in the workcell, we automatically generate contact patches on the part for probing. We then estimate the probing time by considering the different subset of patches and select the subset with the least estimated probing time. The robot motions for part probing consider the potential error in the initial estimate to keep the part safe during the contact-based localization procedure. The probed positions of the contact patches are used to reduce the uncertainty in the part pose. We have characterized the performance of the developed method using both simulation and experimental studies. <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1.1.</sup> DISCLAIMER: Certain commercial equipment instruments, or materials are identified in this paper to foster understanding. Such identification does not imply recommendation or endorsement by National Institute of Standards and Technology. It does not imply that the materials or equipment identified are necessarily the best available for the purpose.