This thesis investigates human performance in relation to various dimensions of 6 degree of freedom (DOF) interfaces, including device resistance, transfer functions, muscles groups and joints, and input display formats. These dimensions are analysed respectively in terms of human proprioception and control feel, mental processing in forming control actions, motor and sensory cortex representation, and the nature of various visual depth cues.
A series of five experiments are presented. Experiment 1 examined isotonic versus isometric resistance modes and position versus rate control transfer functions. A strong interaction was found between the resistance mode and the transfer function: in position control, the isotonic device outperformed the isometric device; whereas in rate control, the isometric device outperformed the isotonic device. Experiments 2 and 3 studied isometric versus elastic devices in rate control. When optimised between two opposing factors, i.e. proprioception and compatibility, the elastic device had performance advantages over the isometric device at the early stage of learning. Experiment 3 also revealed users' control strategies in terms of attentional priority to each degree of freedom. Experiment 4 investigated the effects of different joints and muscle groups on 6 DOF manipulation. The results showed that the participation of fingers significantly improved task performance. Experiment 5 studied the visual representation of users' input control actions. It was found that partial occlusion through semi-transparency in 3D, a rather novel graphic technique, was strongly beneficial.
The major conclusions of the thesis can be summarised briefly
as: (1) The physical properties of a 6 DOF input device should
provide rich feedback so that the user can easily feel her control
actions proprioceptively and thus learn the task quickly. (2)
To the extent possible, fine small muscle groups and joints should
be included in the operation of input devices. (3) The transfer
function used to interface a device with the computer should be
compatible with the physical device. (4) The visual representation
of the user's actions should be designed to allow immediate exteroceptive
feedback and the application of semi-transparency serves this