Gripper Point Contacts for Part Alignment

Mike Tao Zhang and Ken Goldberg 
<2000 ~ 2002> 
ALPHA Lab, University of California, Berkeley. 
 

 Gripper contacts align the part for assembly.

Abstract:

The initial resting pose of many industrial parts differs from the orientation desired for assembly. We show that it is possible to align parts during grasping using a standard parallel-jaw gripper. A solution is an arrangement of four gripper point contacts that will align the part in the vertical plane as the jaws close. Given a n-sided polygonal part and k sample points on each side, we develop an O(n4k4) numerical algorithm to compute a set of solutions or a report that no solution exists. The algorithm combines toppling, accessibility, and form-closure analysis. We have implemented the algorithm and report sensitivity data from physical experiments.

                

Introduction:

“Grippers can be the most design-intensive components of an assembly system”. Although grippers are widely used for automated manufacturing, assembly, and packing, designing gripper jaws is usually ad-hoc and remains a   limiting factor in industrial applications. This paper proposes a new approach based on mechanics and part geometry.

Industrial parts on a flat worksurface will naturally come to rest in one of several stable orientations, but it is often necessary to rotate parts into different orientations for assembly. The figure above illustrates how parts can be aligned using a standard parallel-jaw gripper. The part is initially in stable orientation (a); as the jaws close, the part is passively rotated into orientation (b) for assembly onto the peg.  Given part geometry, we study how to compute the position of four gripper contacts. We illustrate notation in the figure below. Pushing tip A’ and toppling tip A make contact with the part to rotate it from the initial stable orientation to its desired orientation, and then fixturing tips B’ and B make contact with the part, stop its rotation, and securely grasp it. The gripper with four jaw tips is low in cost, footprint, and weight, and can be rapidly reconfigured to handle different parts. 

 

Experiments:

We implemented our jaw contact design algorithm as an application using the Java programming language. Mouse input allows a user to draw a part, define its COM and friction, and select its initial and final orientations; the program then computes and displays the resulting solutions or reports that no solution exists.

To explore robustness to initial conditions, we conducted physical experiments using an AdeptOne industrial robot and a pneumatic parallel-jaw gripper (Mecanotron serial number: 101167). The gripper jaws were designed by the algorithm and manually assembled from aluminum stock. To control the velocity of our pneumatic gripper, we added two air regulators (Wilkerson serial number: R08-01-F000) and a compression spring (Century serial number: C-606).

 

Robot picks up parts from a convey belt using parallel-jaw gripper.

 

The experiment we conducted was to design four-contact jaws for a small lever from a standard videotape (Fuji serial number: 7410161160). Its natural resting pose is as follows:

We have to insert a peg, which is on the video tape case, into the hole of the part for assembly (see the figure below). Therefore, rotation of the part in the vertical plane is required. 
 



We conducted a successful experiment in sequence 1~5 for 50 trials (Click the image below to see a movie, 5M).
 

 

Publications:

  • T. Zhang, G. Smith and K. Goldberg. "Compensatory grasping with the parallel-jaw gripper," in Algorithmic and Computational Robotics: New Directions, edited by B. Donald, K. Lynch, and D. Rus, A K Peters, Ltd., 2001.
  • M. T. Zhang and K. Goldberg. "Gripper point contacts for part alignment," IEEE Transaction on Robotics and Automation (accepted). [PDF]
  • T. Zhang, G. Smith and K. Goldberg. "Compensatory grasping with the parallel-jaw gripper," in 4th International Workshop on Algorithmic Foundations of Robotics, Hanover, NH, 2000.  

Media Coverage:

Acknowledge:

This work was supported in part by the National Science Foundation under DMI-0010069, CDA-9726389 and Presidential Faculty Fellow Award IRI-9553197. Research funding was also provided by Adept Technology, Inc., Ford Motor Co., and California State MICRO Grant 00-032.