[ELECTRON] possibility of short term paid employment

[Gordon JC Pearce]

On Tue, 2010-03-02 at 14:22 +0000, Marion Hersh wrote:

> ·         Investigating options for input/output combinations for the
> deafblind person, including relay switch vibrator/pressure sensor
> combination; pressure sensors and vibro-tactile actuators embedded in
> a smart fabric; miniature vibrators and motors, possibly in
> combination with a relay switch.

The first thing that springs to mind is Morse code, but I might be
biased ;-)

Rather than detect the timing of dots and dashes (which is notoriously
hard to do even if you're sending very accurate Morse), use two contacts
- one for dots, one for dashes.  Normally this would be something like a
paddle key: 

http://en.wikipedia.org/wiki/Telegraph_key#Iambic_.28dual-lever.29_Paddles

In this case, you could use a small blob of something conductive on the
fingertips and thumbs of a glove.  Touch your index finger to your thumb
for a dot, and middle finger to your thumb for a dash.

I'm not sure how well a pager motor would work - would it spin up
quickly enough to give a nice crisp start to a dot?  Stopping it
instantly would be easy using magnetic braking (short across the motor).
Perhaps a small solenoid that would "tap" the back of your hand - back
rather than palm so you can still grip things!

> ·         Investigating input/output options for the communication
> partner, including a miniaturised keyboard, pointer and on-screen
> keyboard, small LCD text display, possibly as part of a mobile
> telephone or PDA, and speech output with text-speech conversion and
> either an earpiece or small loudspeaker. The text display could also
> act as the screen for the screen keyboard.  There would be advantages
> in the development of two versions, one as a self-contained device and
> another which is compatible with mobile phones with Wifi and/or
> Bluetooth.

Decoding Morse and displaying it on a small LCD would be a Small Matter
of Programming; there is already an Arduino library for driving common
"Hitachi"-type LCDs, and decoding Morse.  Entering text may be more
difficult simply because you've got to get a keyboard in there
somewhere.  For initial experiments you could use a plain vanilla PC
keyboard; the serial protocol is well-documented and easy to implement.

You'd need something with a bit of grunt to do wifi or bluetooth.
Likewise, text-to-speech and even more so speech recognition would
probably tax even very powerful microcontrollers so you're probably
looking at something like a Beagle Board.

> ·         Writing software, including for communication between
> communication partner and deafblind person and for input to and output
> from mobile phone and other input/output devices.        

If you stick to nice simple class-compliant USB interfaces like
usb-serial devices, you'll avoid a lot of hassle in the long run even if
it means you need a "clever" serial protocol.  Software is easy.

> ·         Investigating durable materials for the glove.

The difficulty with wearable computing is that you've got to take the
electronics out to wash the clothes!  Even if you just leave the cables
in place, washing will potentially damage the connectors and corrode
wiring inside the insulation.  You don't want that.  Maybe some sort of
durable breathable material that you could pull over a washable glove?

Obviously it doesn't have to be a glove - you could use a suck-blow
switch for tetraplegic people.  You could use a footswitch for someone
who has no use of their arms.  The possibilities are limited only by the
number of muscles you can use to move a contact.  Detecting nerve
impulses directly is very hard, but possible.

Gordon MM0YEQ