The text below is several years old and in the interval smart phones have improved so much that they are now pretty close to what most people mean  by "wearable computer" --- except for their little keyboards.  Neither touchscreen boards nor thumb boards are good enough to let you work all day on your smart phone as you could on your desktop computer.  Data input, the I in I/O, remains the bottleneck for mobile devices.

The What & Why Of Wearable Computers

Most people, on first hearing "wearable computer," probably thought "who needs it," closely followed by "how goofy looking would that be."  Nevertheless a huge market will develop as portable computing hardware becomes more portable, less obtrusive and takes on more of the functionality of desktops.  

In some circles, mobile means roughly "works on a moving vehicle" and "portable" means "can be carried by a pedestrian."  I'd rather say here that both refer to the continuum of motions and postures allowed to the user.  Thus a device which requires the user to be seated may be considered less portable than one for which the seated posture is optional.  The more things you can't do while using it, the less portable it is.  

A (simple, non-smart) cell phone is more portable than a laptop computer because you can operate your cell phone with one hand, standing, walking or sprawled upside down on a couch and that's certainly not the case with your laptop computer or even your PDA.  This isn't due to any limitation of the internal electronics.  The problem is the I/O peripherals.

The conventional screen and keyboard have a natural scale determined by the dimensions of the human eye and hand.  Shrink the keyboard or the screen too much and they become unusable.  Many of today's devices are near the limits.  And since the screen and keyboard can't be much smaller there's less pressure to shrink other parts.  The way forward, of course, is to replace the conventional screen and keyboard with designs that are not subject to the same limitations.

In doing so we cannot avoid the all-important issue of social acceptance --- the geek factor.  Few really want to look like a Hollywood cyborg.  Besides being usable in any posture, the ideal wearable computer would be invisible, at least while stowed.  The geekier a wearable looks the more social situations it will be denied.  The wearables available today won't soon be appearing anywhere people are trying to impress one another socially.  But that still leaves immediate industrial, military, academic and recreational markets --- if functionality more comparable to desktops can be achieved.

A widely shared vision of wearables is that the main output device should be a small screen or, better, a virtual display that can overlay rather than obstruct the user's view of his real environment.  The main input device should be a one-hand keyboard and everything should be as small as possible.

Technical Details Of The One-hand, "Chordite" Keyboard

Briefly, the big ideas are

  1. Relieve the typing fingers of any role in supporting the keyboard. Do this with a frame held between the thumb and palm rather than using a wrist strap.

  2. Place the keys so that the typing fingers remain near their relaxed positions.

  3. Use two keys per finger. Press one with the distal phalanx and the other with the middle phalanx.

Number three is, I believe, novel. Until you try it's not obvious that it works.

The human hand is a marvel of nature and has plenty of bandwith and endurance for prolonged data entry if you interface with it on its own terms.  Of course to be fast, the finger motions should be small and light.  The hand also expects appropriate tactile feedback.

Not surprisingly, fit matters.  Fit is important with almost anything that touches the body.  Qwerty keyboards are a rare exception --- they don't need to be adjusted for hand size because the user can compensate for poorly placed keys by moving the entire hand relative to the keyboard, adjusting keystroke by keystroke.  This degree of freedom is not available when the operating hand also supports the keyboard.  Consequently the fit must be more precise for portable one-handers.  I currently believe that one-size-fits-all one-handers are possible but will likely be table mounted.   Most people, IMHO, will prefer units adjusted to fit their particular finger dimensions.  Thus one-hand keyboards may come to be considered personal effects like eyeglasses or hats.

Experimentation has convinced me that the optimal switches for the stealthy keyboard would be soft (30 gf or less, like those on a standard qwerty board), have a short throw (a millimeter or less, like a metal dome tactile switch), life expectancy in the tens of millions (like a qwerty's) and preferably a tactile snap. So far I haven't found any keys with this particular combination but it's clear there's nothing undoable about it.  I'm just not ready to order 50,000 switches to get exactly what I want.

The best switches I could find for my prototypes are of the type called "micro switches" or "basic switches" or "basic snap switches."  The hinged levers aren't as desirable as a straight-down push action would be but suffice.

Among the biggest shortcomings of my prototypes are that they are not quickly adjustable and have no miniature joystick.  A commercial product definitely requires both these features.  A commercial product might also be wireless, have a display screen, etc.