CS789, Spring 2005

Activity instructions, Week 5

Use a 3D input/output device not connected to a computer.

You are certainly very familiar with standard input and output devices like the mouse, keyboard, screen and speaker. They are either logically one dimensional -- keyboard, speaker, teletype -- or logically two dimensional -- mouse, screen. Many less common devices, the trackball for example, are similarly logically two-dimensional. Thus, we should not be surprised if tasks that are inherently one-dimensional, like text processing, and tasks that are inherently two-dimensional, like illustration or page layout, are pretty well-served by existing input/output devices. We should also not be surprised if tasks that are inherently three-dimensional, like industrial design, are poorly served by existing devices.

A common sign that there are unsolved problems in a particular group of tasks is substantial effort at innovation in that general area, start-up companies, new products, academic research. By that measure the state-of-the-art in three-dimensional input devices leaves something to be desired. Typical applications that obviously benefit from such technology include complex visualization, industrial design and 3D computer animation. We see a variety of solutions to the three-dimensional input problem. Some are listed below, in roughly increasing exoticism. (There should be references for each, but life is short.)

• extra input modes on conventional input devices, the 3D trackball,
• multiple views of the designed object, perspective, orthogonal,
• manipulable perspective views,
• 3D generalizations of conventional input devices, the bat, the spaceball
• binocular displays,
• head-tracking displays,
• fishbowl virtual reality,
• immersive virtual reality.

In addition, methods have been evolved for some design activities that can be approached less interactively, specifying shape using geometric primitives, for example. Even so, much three dimensional computer-aided design is still performed in the following very kludgy way. A designer makes a shape using modelling clay; the shape is hand or laser scanned to get it into the computer; the computerized shape is painstakingly hand-editted; the result drives whatever computer-mediated processes follow: machining, mold-cutting, part layout, wire-routing, or whatever. The goal of this week's activity is to use modelling clay to do three-dimensional design, consider what unique properties it might have, and think about how they might be incorporated into a 3D input/output device.

1. I will give each of you a some modelling clay in class. Spend half an hour shaping it into some object that interests you, a fender, a toaster, a hand, a molecule, or whatever.
2. While shaping the clay think about how the clay gives you tactile and visual feedback showing how it is changing in response to your actions. Think also about the coupling between the clay itself and your fingers: how does it change when you push it?
3. Also think about how the clay changes as you work it. Does this play a role in making clay a desirable medium?
4. Think about conventional ways of doing this task with a mouse and screen. What is missing from them, that is present in modelling clay? Which missing things are actually important?
5. How would you design a device that would be better for 3D than the mouse? What properties should it have? How would you translate them into technology, both hardware and software? Consider temporal factors as you do this.
6. Make a business plan; find a venture capitalist; become a millionaire.

Return to:

• Bill Cowan's home page.
• Bill Cowan's teaching page.
• Bill Cowan's CS789 page.
• Bill Cowan's CS789 page for Spring 2005.
• Bill Cowan's CS789 page for Spring 2005 : Activities