First, an anecdote. Eugene Fubini was visiting Livermore in perhaps 1959 and we were describing to him the computer graphics work that we were doing, including stereo vision. He said that he wanted to reach out and feel the molecules that we were displaying. I thought that he was kidding at the time but I realized later that he may not have been. Already there were “waldos” which provided two way force and positioning information between “gloves” and remote manipulators. Some of these were purely mechanical but some were electrical if not electronic.
Recently I wrote a program that I had wanted to write in the early 60’s to perform 3D elasto-dynamics. There is a simple demo version here of a rubber rod which you can interact with from the keyboard. It would be fun to have a handle for the rod in real space, with which you could grasp the virtual rod and move it around and feel its inertia and vibrations. I have described a six axis stress gauge which would be one element. A six axis robot arm would be the other more complex but also more standard part of the system. One facet of the gauge would be attached to the arm and the other attached to a graspable handle.
Of course I should first explore off the shelf waldo gloves, if they come disconnected from the manipulator.
All of this is far from feeling something in cyberspace. It would be like feeling something while wearing iron gloves. I cannot yet imagine how to do a virtual texture, short of nerve transducers.
The pedipulator is of the same ilk; it deals with actuation by people and force feedback to them.
Today (2012) such technology is referred to as “haptics”. I am not much impressed with the projects mentioned there except for the virtual Haptic Back Project which aspires to provide realistic force feedback to the user as immediate response to his motions—realistic resistance to user movement. (This robot hand is also impressive too but not what I would refer to as haptics.)