Evaluation of an automated three-dimensional compensation algorithm for visual-display misalignment and effects of display formats in three-dimensional telerobor manipulation
In teleoperation environments, decision-making can be performed by a combination of knowledge-based autonomous procedures, sensor-based autonomous procedures, and/or the human operator. Humans can easily adapt to unpredictability in task environments, due to their superior problem solving skills and perceptual abilities. Therefore, using a human operator to make decisions is beneficial to the manual control of telerobot in real environments.
When a robot is manually controlled in teleoperation, the control input of the operator is transmitted to the robot, and video cameras send visual feedback of the state of the robot to the operator. In this manner, the operator is engaged in the dynamic control of the robot. Some characteristics of this control have disadvantages. The visual feedback of remote manipulation by the video cameras requires a very high communication bandwidth to transmit the video signal. A small communication delay in control feedback deteriorates the teleoperation performance of the operator. The other disadvantage is due to spatial perturbations including depth problems and visual-display incompatibilities. Those perturbations can be reduced by training operators or using graphical aids. However, it usually takes a long time to get satisfying results through training, and most of the time it is very hard for operators to reach a satisfactory level of performance.
The teleoperation visual system needs to provide sufficient visual information to allow various tasks to be accomplished. Understanding the relationship of the manipulator to some fixed reference plane is the basis for spatial orientation, and displaying control disorientation can result in degradation of operator performance as well as damage or loss of the manipulator. When multiple cameras or dynamically moving cameras are used in a manual operation, depending on the camera view angle, the axes of the manipulator are not aligned with the controller axes. This misalignment causes display-control incompatibility. Under the incompatibility conditions, the performance of the operator might be lower than the performance in compatibility conditions. In this research, 3-D automatic compensation method for visual-display compatibility was tested to reduce visual-display incompatibilities. Three different display formats with the 3-D compensation method were tested in telerobotic tracking simulation environments.
The 3-D automatic compensation method can be applied to the display-control incompatibility conditions to reduce incompatibility. There is no need to change hardware settings for integrating the 3-D compensation method. In experiment I, the 3-D compensation method was integrated with the single display format. In experiment E, the 3-D compensation method was integrated with the three display formats and a visual mode. When the compensation method was used, performance was superior to the performance conditions in which the compensation method was not used. In addition, the single monitor with the 3-D compensation method saves cost by using a monitor and a low speed network connection as compared to using the multiple monitors. In each view, the operator works under display-control compatibility conditions, so that the operator freely select a good view without considering the display-control compatibility.