Intelligent Visual Servoing for Robotic Control
Abstract
The Department of Defense Critical Technologies Plan (March 1990) targets machine intelligence and robotics as one of 20 "technologies with great promise of ensuring the long-term superiority of United States weapon systems." Robotic systems and technology are being developed and applied to the important requirement of keeping mobile weapons platforms such as tanks and artillery supplied with ammunition in combat situations. This application of robotic technology can both increase the efficiency of the ammunition handling and supply operations and, at the same time, reduce the exposure of military personnel to hazardous environments and missions. RedZone Robotics, Inc. has completed a Phase I SBIR that addresses the needs of DOD SBIR solicitation A94-094, "Intelligent Sensor Based Robotic Control System Technology". This project has resulted in a design for intelligent vision-based servoing control of robotic and telerobotic manipulators that will enhance the capabilities, adaptability, robustness, and autonomy of existing robotic controllers for tasks that involve object tracking and manipulator positioning. This technology is modular in structure so that it can be added to existing robot control systems or integrated into new robotic and telerobotic control systems. The visual servoing system incorporates novel video feedback algorithms, using a single camera mounted on the robot, to actively position a manipulator relative to target objects. Rather than relying on precise, pre-movement kinematic modeling, the technique continuously measures target position in a video image while the robot moves. Visual servoing relaxes conventional requirements for object tracking and manipulator positioning and will allow less expensive, faster setup of new and existing robot systems. The ultimate result of this research will be new intelligent visual servocontrol commercial products for semi-autonomous robotic or telerobotic operations.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 17, 1995
- Accession Number
- ADA365931
Entities
People
- Alan D. Berger