Original Date: 01/27/1997
Revision Date: 01/18/2007

Robots typically are programmed through either a teach mode or by manually correcting programs that have been downloaded from a model-based programming system. Both methods reduce flexibility and require extensive programming time. In industry, most robotic applications use the pick-and-place methods which lack part recognition or collision avoidance capabilities. Lawrence Livermore National Laboratory (LLNL) has developed a telerobotic system which increases flexibility and autonomous capabilities, reduces programming time, and offers a 3-D shape recognition system.

The alliance of robotics and the stereo 3-D surface imaging system has made this system very unique. The telerobotic system allows the robot to perform random-part recognition, controlled-force compliance, and real-time path planning. With a force reflecting hand controller, the robot’s end effector can be programmed to move to a part or surface and either grasp or avoid contact with it. By using the 3-D imaging system, the system enables the robot to visualize a part and its orientation and then quickly move it to its designation.

LLNL’s telerobotic system offers major advantages over existing robotic systems. Advantages include an easy-to- use control system for teaching and operating the system; a collision avoidance and force compliance capability to keep the robot from damaging itself or other equipment; programmable operations which use force compliance, 3-D vision, and a graphic model to describe a work cell instead of teaching the robot to move to set positions; and automatically-planned, collision-free paths for performing new tasks. The 3-D imaging system, an integral part of the telerobotic system, is a low end unit which does not require a laser as the active light source, and can be configured for a PC host for less than $10 thousand.

For more information see the Point of Contact for this survey.