Situations in which the human operator must control three translational degrees of freedom are common in, for example, control of robot arms, remotely controlled cameras, and tele-surgery. Often, the operator has no direct view on the controlled object, but receives information on the motions of the object via a camera-monitor system or a simplified graphical display. In the design of such systems, it is important that the mapping of the motions directions of the input device and the controlled object areompatible for the human operator. A compatible relation will lead to faster reaction times, faster learning and less errors. For this relation, two mapping principles may be distinguished. The first is a mapping in which the motions of the input device and the controlled object are always parallel in 3D space (spatial-motion mapping). The second principle is a mapping in which the motions are parallel by comparison with the ground planes they are located or displayed in (reference-plane mapping). These ground planes often are not parallel, for example a monitor placed on the operator table in which the control is located. When the ground plane of the device and the object are not parallel, both principles will lead to a different mapping. The literature on tasks with two degrees of freedom shows that reference-plane mapping is the more compatible mapping principle. However, the question is whether these results are valid for 3D tasks as well. Research on control-display mapping for three translational degrees of freedom is scarce. A previous experi-ment at the TNO Human Factors Research Institute showed an advantage of reference-plane mapping. Nonetheless, there are indications that this advantage is caused by the restrained 3D visual depth cues, which encouraged subjects to experience the task as a 2D task with a third dimension added. The present experiment uses the same positioning task, but with a display including high quality depth cues. This may be seen as a simulation of indirect viewing via camera-monitor systems, instead of a simplified graphical representation. The main result found was the beneficial effect of spatial-motion mapping. It was suggested that this effect differs from the effects found in the former experiment, because of the presence of high-quality visual depth cues. These cues may have encouraged subjects to experience the task as a real 3D task. This result may have serious implications for the designers of systems with three translatioional degrees of freedom. The compatible relation between motions of the control device and the controlled object may be largely dependent on the availability of visual depth cues in the display. When these cues are not available, operators may prefer a reference-plane mapping. But when these cues are available, the spatial-motion principle is preferred, which leads to a different mapping.
Gesuggereerd wordt dat het aanbieden van drie dimensionale visuele cues invloed heeft op de gepreffereerde mapping van bewegingen van besturingsmiddelen en cursor als deze zich niet in hetzelfde vlak bevinden.