|Abstract ó Nuclear facility dismantlement tasks include disassembly of process equipment, cutting pipe, size reduction of equipment, transport of materials, and decontamination of floors, walls, and remaining equipment. Plans for using robots to perform these tasks specify direct teleoperation in the near-term with a transition to more autonomous robots as the technology becomes available. As this control progresses from direct teleoperation to autonomous robots, the sophistication of the human intervention must correspondingly increase. This paper discusses different levels of interaction between a human operator and a remote redundant robot. The discussion includes development, implementation, and test results for five different modes of teleoperation. Two of these modes include a degree of computer-based decision making. Results show that the computer is capable of making very fast decisions when confronted with complex kinematic problems. The difficulty lies primarily in the computerís need for information about the location of obstacles in the robotís workplace.
Introduction - A spectrum of tasks in unstructured environments characterizes the Decontamination and Dismantlement (D&D) mission. Telemanipulators represent a technology for accomplishing a portion of the missionís task spectrum. Telemanipulators allow the human to project manipulative capabilities into remote hazardous environments. Giving these telemanipulators extra kinematic degrees of freedom (kinematic redundancy) enables them to perform a wider range of tasks, thus further amortizing costs. Developed by the United States Department of Energy (DOE), the Dual Arm Work Module (DAWM) is an example of a telemanipulator with kinematic redundancy. The DAWM has 17 Degrees Of Freedom (DOF) arranged in 2 serial chains each having 8 independent DOF and sharing 1 common center rotational joint. Red Zone Robotics manufactured and delivered the 5 DOF base unit that includes the common rotational joint and the next two joints in both chains. Schilling Titan II manipulators form the last six DOF for each chain. With two arms and 17 DOF, the system has 5 degrees of kinematic redundancy.
This kinematic redundancy gives the system the versatility to perform a very wide range of tasks. Furthermore, the redundancy typically affords a number of different kinematic options for performing each task. In other words, the robot can move its body while holding its End-EFfectors (EEF) at constant locations. These motions represent sets of configuration options. The operator (or the robotís own internal decision-making algorithms) can choose an option best suited for a given task. A skilled operator will choose an option based on visual feedback and experience with important system parameters. These parameters represent operational performance criteria: joint motion limits - travel, velocity, acceleration, torque dual arm criteria - relative load, energy, compliance obstacle avoidance - model for system and environment task criteria - force, deformation, dexterity. Machine-based decision-making algorithms must match or surpass this level of sophistication before they can begin to share control with human operators.