Robotics and Automation Expert
page 1 page 3 page 4 page 5 page 6 page 7

Manual Control of Teleoperated Redundant Robots (page 2)

This paper discusses five control modes developed specifically for the DAWM, though they apply in general to all redundant telemanipulators. These modes are listed below in an order of increasing sophistication.

Direct Joint - Drive each joint independently
Decoupled Cartesian - A six-joint subset in Cartesian, and extra joints independently
Self-Motion - Drive individual joints while maintaining hand position
Advisor - Computer suggests configuration based on task specifications
Coupled Cartesian - Computer automatically controls all joints based on hand controller input and task specifications

These control modes represent steps in the hierarchy of teleoperation. Draper describes a continuum in the sophistication of the operator’s interaction with the machine during teleoperation) In an order of increasing sophistication, the types of control in the continuum follow as: manual control, intelligent assistance, shared control, traded control, and supervisory control. The first three modes from described above are examples of manual control. The advisor mode is an intelligent assistant and the coupled- Cartesian mode is in the category of shared control. 

Manual Control - The lowest three control modes fall into the category of manual control. Though these control modes employ very sophisticated electronics and computer algorithms, there is no machine-based decision-making involved as the operator controls the robot’s extra joints. Thus, the direct joint, decoupled-Cartesian, and self-motion modes of teleoperation fall into the category of manual control.

Direct joint control is the simplest mode. In this mode the operator drives each of the robot’s joints independently. Precisely controlling the location of the robot’s end-effector is very difficult in this mode. The mode is, however, very useful for extricating the robot from difficult configurations. These difficult configurations may involve joint travel limits, singularities, and workspace boundaries. In the joint control mode, the operator can simply jog each joint until the robot is in a more favorable configuration. After that, the operator can switch the control mode to a more sophisticated level of interaction.