6-DOF Haptic Rendering


Abstract: We present an algorithm for haptic display of moderately complex polygonal models with a six degree of freedom (DOF) force feedback device. We make use of incremental algorithms for contact determination between convex primitives. The resulting contact information is used for calculating the restoring forces and torques and thereby used to generate a sense of virtual touch. To speed up the computation, our approach exploits a combination of geometric locality, temporal coherence, and predictive methods to compute object-object contacts at kHz rates. The algorithm has been implemented and interfaced with a 6-DOF PHANToM Premium 1.5. We demonstrate its performance on force display of the mechanical interaction between moderately complex geometric structures that can be decomposed into convex primitives.

Paper Six Degrees-of-Freedom Haptic Display of Polygonal Models (PDF, 78 KB)
Arthur Gregory, Ajith Mascarenhas, Stephen Ehmann, Ming Lin, and Dinesh Manocha
Color Plate
Color Plate (Appears with paper. PDF, 153 KB)

Polygonal Datasets
 Gear Demo: 
The gears are modeled and positioned in such a manner that the user can turn one gear with the other in either direction. The teeth are not so tightly interlocking that there is always a collision. In the demonstration shown, the gears contain a few hundred polygons and about 20 convex primitives. The user can attach the haptic probe to either gear in two different modes. Clicking the button towards the rim of the gear is like inserting a rod into a bicycle wheel. The probe's position remains at a fixed radius from the center of the gear and its orientation is fixed so as to rotate with the gear at that radius. If the user clicks around the center of a gear, the probe effectively becomes the gear. Its position is fixed, and it is only allowed to rotate about one axis. In this mode, the user turns the gears and feels their interactions entirely through the torque of the probe handle. 
 Peg-in-the-hole Demo:
In this scenario, the user attaches the probe to a rectangular peg and attempts to insert it into a rectangular hole. Often one or two pairs of the parallel faces are the pair(s) of the closest features and may be in contact. If any type of  sampling technique is used, the number of contact points would be very high, since nearly all faces of the peg are in close proximity with the walls of the hole. Collision detection and contact determination for such a seemingly simple scene are actually rather difficult, due to its contact configuration and geometric robustness problem. 
 Dice and shaker: We have a dynamic scene with objects moving under the influence of gravity. There are three dice in a box-like shaker. The user can grab the shaker and influence the motion of the dice. Gravity results in the user feeling continuous force and torque and the collisions between the shaker and the dice result in impulse force and torque which are felt as small blips.
 Plank and weights: The user can manipulate a plank and a pair of cubes with unequal weights.
Dynamic Scene
Dynamic scene of multiple moving objects: In this setup, there are four cubes, four spheres (320 faces each), four ellipsoids (320 faces each) and a stick like block. The user can grab any object and move about interacting with the other objects.
Volume Datasets
Electric Charges
Force-field due to two electric charges: The force-field and streamlines were generated using Greg Turk's streamlines program. The user can move a rigid-body through the field using the 6-DOF device and feel force and torque exerted by the field on the rigid-body. In the scene there are two charges of opposite polarity with the charge on the right acting as a source and the charge on the left acting as the sink. 
Force-field due to uniform flow inside and around a cylinder: The scenario shows ideal fluid flow in and around a cylinder. The user feels strong unstable forces at the singularities inside the cylinder. 
Utah Head
The Utah head Data-set:  The image shows the potential values on the head data-set. The head volume is tetrahedrized. There are three electro-static charges placed inside the volume and the electric field vectors and the potential scalars are given at each vertex of the tetrahedrization. The user can move a rigid-body like a cube in the volume and feel the force and torque exerted on the rigid-body.

Principle Investigators
Ming C. Lin
Dinesh Manocha
Project Members
Stephen A. Ehmann
Ajith Mascarenhas
Past Project Members
Arthur D. Gregory

Related Projects
 UNC Research Group On Modeling, Physically-Based Simulation And Applications Geometric and Solid Modeling
 The nanoManipulator Graphics Groups at UNC

Funding Agencies
 Army Research Office
 National Science Foundation
For more Information, contact geom@cs.unc.edu

Created by Ajith Mascarenhas
Last Content Update : Aug 19, 2000
Last Content Review : Aug 19, 2000

Copyright 1999. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the author.

This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.