Instrumentation is designed to be used with a "chorded
keyboard". This page is a further attempt
to design a keyboard that is compact, wearable
and rugged. These diagrams are really just preliminary sketches, so
the keys and slots may not actually line up exactly and the
dimensions may not be optimal for an actual hardware implementation.
Here is an exploded view of a right hand keyboard. The sides have
been flattened out to show the matching slots and keys (for locking
the right and left hand keyboards together), the strap connectors
and the thumb buttons. Normally, the upper, lower and right sides
(as shown below) would be permanently folded back to form a rigid
box. Only the left side would unfold to create a full length
keyboard. The slots are shown with dashed lines because they are on
the inside (the other side in this view) of the deep sides.
Any resemblance to a concert grand piano is purely causal
An optional strap would secure the keyboard to the hand so that
chording would not require any external support (like a qwerty
keyboard does). The strap will connect at the four corners of the
keypad as a thick 'X'. Hopefully this will provide reasonable
stability.
I have not shown a power switch or the connectors for recharging,
display, external memory, etc. This design is much too basic to
worry about the placement of internal or external hardware. Physical
mockups will be required to flesh out the most efficient and
comfortable arrangement of any necessary additional connectors and
controls.
I placed the buttons by using my own fingers as a
guide. The actual keypads would probably have oval buttons to
accommodate a wider range of finger lengths. The second thumb
button would be used to provide a command mode for controlling
functions outside the range of data entry (like making a phone
call, or ejecting a memory chip).
This design specifies capacitive buttons because they have no
moving parts. This would also minimize pressure on the locking
hinges because no force would be required to activate the buttons.
This design lacks the ability to provide vibrational haptic
feedback, but that option could be offered on a larger, heaver,
less robust and more expensive model.
The keypad would be unfolded for use and then re-folded for
storage. It might not be desirable to unfold the keypad
until it was fully flat. A keypad that only unfolds to roughly
ninety degrees would eliminate one hinge and could probably be
made stronger, more compact and more stable. Also, bending the
fingers allows the finger tips to align. This would mean that the
buttons could be placed closer together and a single keyboard
could accomidate a wider range of hand sizes and finger lengths.
I'm sure other improvements would become obvious if a three
dimensional prototype was created. An empty plastic cassette tape
box would be an adequate platform for experimentation, but I'm
still working on the vocabulary, software and educational portions
of the Instrumentation project, so I haven't progressed beyond
drawing these inaccurate pictures.
Here are the folded right and left hand keyboards shown together.
The dotted (outer) shallow sides are the hinges. 'Shallow' refers
to the depth of the side rather that its width or length. This is
clearer in the previous flattened view.
The keypad numbers match the octants in the glyph
below. The two sets of numbers are reversed (seven is on the
bottom above and on top below) because I was holding my hand up to
the screen when I drew the first keyboard view. In either case,
seven is the pointer finger of your right hand.
Having two keyboards would increase the amount of memory,
connector space (for external appliances) and processing power.
For storage, the keyboards would be rotated so that the keys on
the outside of the shallow sides would mate with the slots on the
inside of the deep sides. The deep sides would completely cover
the shallow sides and protect the thumb buttons and any connectors
for external hardware. In this design the strap connectors remain
extended when the keyboards are locked together, but there may be
a better way of connecting the 'X' straps.
When closed, there would be a quarter inch gap between the keypad
surfaces and a half inch gap beyond the ends of the keypad
surfaces. These gaps would provide protection and storage space
for cables and straps. The keyboards could be cordless (or use the
wearer's
body as a data conduit), but a cable would provide the most
stable and secure communication with the greatest bandwidth.
Using the keyboard
The octant numbers on the glyph (above) match the keypad numbers on
the keypad diagram (farther above). You can get a better idea of how
a glyph works by playing with the game.
A glyph can entered as four or less 'chords'. A chord is played by
pressing multiple buttons simultaneously. The typist would press all
buttons that correspond to visible (i.e. black or blue) elements in
a given glyph layer according to the numbers in the figures above. A
chord need not be entered if that layer is not present. The chords
could actually be entered in any order because the thumb
combinations make them unique.
The first chord would enter the inner
spokes.
The typist would press the appropriate finger buttons.
The second chord would enter the inner laths.
The typist would press the left thumb button along with the
appropriate finger buttons.
The third chord would enter the outer spokes.
The typist would press the right thumb button along with the
appropriate finger buttons.
The fourth chord would enter the outer laths.
The typist would press both thumb buttons along with the
appropriate finger buttons.
The hub would be plugged unless the typist pressed both thumb
buttons without pressing any finger buttons.
A solo right thumb press would advance focus to the next
glyph.
A solo left thumb press would return focus to the previous
glyph.
pressing any (non-thumb) button a second time (before entering
the next layer) would toggle an (incorrect) element to the
opposite state within the current layer.
It should also be possible to reverse these left and right
actions for people who are left handed or people who normally
read 'right to left'.
The "User Interface" section gives more
examples of 'chording'.
This design does not address the glyph display hardware. A separate
display could be a compact or full sized screen or a heads-up
display using a pair of 'active
goggles'. It might be possible to carry a small screen in the
gap between the two closed keypads, but the gap would probably need
to be widened and the screen would require some sort of support to
be usable.
Caution: The
following has more "blue sky" than usual
I believe that the ultimate expression of this design would
eliminate the keys as a physical objects. As long as the finger
tips can be tracked, normal input should not require
touching anything. The system could be activated by an unusual
movement (touch left thumb to left ring and right thumb to right
pinky [twice], try it!).
The usual argument against this sort of thing is the lack of
positive feedback. I am sure that this will increase the
learning difficulty, but it will not make the skill
impossible to learn.
I think that the advantages of "Hands Free Keying" (HFK: It is
not just an acronym.*) will
outweigh the steeper learning curve for those capable of mastering
it. Visual feedback via the glyph and aural feedback via the
'active' goggles should ease the learning slope (the system must
have some physical presence even if it is just a chip in your
brain).
The fingers wouldn't have any fixed positions, moving a finger
towards the palm would represent one, moving away from the palm
would represent zero and not moving a finger at all would repeat
the previous value. Thumb movement will indicate which chord is
being entered. The glyph for the term 'Instrumentation' (#FF FF FF
FF is the name of this language) would be entered by clenching all
eight fingers then clenching the right thumb then clenching the
left thumb and finally clenching both thumbs (the thumbs never
"enter a zero" so spurious 'keystrokes' shouldn't be a
problem).
Motion
capture technology is common enough now that the hardware
should not be too very difficult to obtain. The current "lack of
sensitivity" would actually be desirable for most users because it
would be more likely to ignore involuntary finger twitches when
entering glyphs.
The location of the sensors could be a problem with a "line of
sight" system such as a laser, but a RFID
system (with passive RFID tags lacquered onto your finger nails)
should be reasonably robust.
"Not to scale"
Such a system would require two sensors. Each sensor would need to
be able to detect an independent range value for each digit. A 3D
location cannot be determined without three sensors, but I do not
believe that this would be a problem because we would be tracking
motion and not position.
If the sensors were mounted at the wrists, each sensor would
monitor a single hand and the orientation of the palms would be
irrelevant. If the sensors were mounted at the temples of a "Head
Mounted Device" (goggles), the differential ranges of all ten
digits would have to be compared to determine the orientations of
the palms. The second method is less robust, but I believe that it
is still possible.
This would eliminate the problem of answering a text when you are
holding cups or other non-squirmy non-metallic things. As long as
you can wiggle your thumb tips and finger tips meaningfully you
can communicate and control your environment.
* It is also not a lot of other
things, things such as 'real'.
