----- Original Message -----
Sent: Monday, April 12, 1999 11:58
AM
Subject: Accu-Touch Design
To all interested parties;
We have now completed the first
phase of the Accu-Touch redesign (design specification). I have attached
a coy of this design to this email. Please review and email me your
comments.
Here is a brief overview of the major points of the
design:
- Weigh less than 15 lbs..
- Data is stored on a PCMCIA (PCCard) memory card
similar to the Accu-Vote but with much more capacity. Therefor the election
that the Accu-Touch processes is based on the memory card inserted (again
like the Accu-Vote).
- All connections, except the smart card
reader, will be behind locked panels.
- Internal battery backed up system good for at least 1 hour hopefully up
to 8 hours (dependant on cost & weight)
- Has two PCMCIA connectors, one for the 'memory' card the other for
expansion such as modem.
- Slim ergonomic enclosure
- Has Infrared port for communicating with
Pollbook
- Has smart card reader
- Headphone & keyboard connector for VIPs units
1. Data must be
stored on more than just one media. We have had some field
experience with this problem. The fact that we have had a disk on
chip has saved our bacon in Tarrent County, Shelby County, and in Gaston
County. Since we do not have paper ballots to reconstruct the
votes casts as the AccuVote unit has if its memory card fails, The Accutouch
will be at the mercy of that one memory card without a
backup.
2. The smart card
reader, dialup modem, and Ethernet connections should be built in to the
unit. Since the addition of these options are typically an extremely low
cost item at the chip level integration to the new mother board, it just makes
since to include these at the time of design. Also a prerequisite for
the engineering team is to design the interfaces so that the actual driver
chips for the above options can be replaced to alternative vendors with some
minor strapping changes on the mother boards and by swapping out the chips
from sockets.
3.
I feel that the PCMCIA should handle an additional flash memory
or hardcard PC card to handle expanded database requirements for early
voting jurisdictions that will require all ballots for all precincts
to be available on each voting station. I believe that without this
functionality our unit will most likely be precinct specific only.
There are a few items I really would like some feed
back on. They are:
Two screen options:
We are currently considering supporting an 10 inch and a 15 inch LCD.
The 10 inch is about $400, our cost, less expensive therefor approx. $1200
list price difference. Is this a worthwhile option? Any ideas as to
what percentage 10" verse 15" system we will sell. Should we consider
supporting a larger screen 19 inch? How much more could we sell this for
($100, $500, $1000, $2000) ?
If we design the unit to the maximum size
screen first then the smaller screens can be adapted with same foot print
just by adding a fill in panel, thus keeping the same look and ergonomics,
but reducing the cost of the panel and being competitive. The
solution for the cabling issue is to copy our present design idea of
incorporation a transition interface between the LCD and touch panels and
the mother board. If we make sure that whatever TFT and Touch driver
chipset we put on the mother board supports multiple sizes of touch
screens and TFT panels. I know that ELO touch systems and Samsung
TFT panels are available in many different sizes. Also, if we need
to change driver chips, the engineering must be in place to allow us to
change these with strapping options and chip replacements as I mentioned
before. We do this today with standard PC mother boards when
changing from an Intel processor to a Cyrix processor just by changing the
chip and some straps. I feel that 15" should be the largest screen
we should support as a built in. All other should be supported as a
external add on via external D-sub 15 pin SVGA connector for the
video, and a USB port for the touch. If we do this then we can
maintain a window open for us to accommodate future special
requirements. Again buy eliminating the internal panel we can
engineer and panel replacement that (a.) maintains the look of the system
and also (b.) provide the external ports needed to interface the
external monitor and touch device.
Internal
printer
Should the printer be
internal or external? The disadvantage of an internal printer is
that is places some serious restrictions on the ergonomic design of the
enclosure due to paper roll size. It also impacts the power supply
requirements since printers consume a fair amount of power. The
advantage is cost. If most units need a printer then the internal
printer is the way to go. So what percent of units are expected to
need a printer?
I see the permanent need of an internal
printer just as we need it for the Accuvote. A possible solution to
the power supply and battery issue is that during power outages we may not
be able to print. Make sure that the power supply only
backup critical power for voting and memory protection. Making
the power feed to the printer a none backup feed should do the
trick. For the ergonomics concern just take a look at
today's small and compact hand held calculators in Office Max.
We just need to open our minds to lighter, cheaper, and alternative
printing conventions, since the final result is to print a paper backup of
a file that we can reprint on demand at anytime.
Landscape verse portrait screen orientation
The expected 'normal' operation of the unit will be
in portrait mode (higher than wider) but we have had requests for
'landscape' mode. For those places that want landscape would the
current 15 inch Accu-touch screen work or do we need an even larger
screen? How many units potentially will want the landscape screen (100,
500, 1000, 5000, etc) ? How much more will they pay?
My thoughts on this is that if
anyone needs Landscape mode it is probable due to Row Voting requirements
and if this is the case I feel that the smallest should be a 17" Landscape
panel or monitor. As described above if you remove the internal
panel and replace it with a mockup panel that serves and a platform to the
external monitor or panel with the necessary connections this will be the
way to go.
My final thoughts, marketing and sales
should be careful of adding functions and requirements to the products just
because a customer wants it. The response to a inquirer on a added
feature or functionally should be, " Engineering must review this to
insure capabilities". This will provide Global a way to place a price
and a value to this option. Even if you decide to discount it later, at
least the customer will know that it was not part of the basic unit and
it was not free . We must have a basic unit that everyone agrees to (
i.e.. 1 hour battery vs. 8 hour battery... Who is right?). We must
not redesign the unit on every sale. Leaving some options for
the future just leaves the door open for upgrades (for a price) in the future
and a continuing capability of forcing your customer base to keep up with the
Jones when they see a new product three, four, or five years down the
road.
Also, we must sell what we have now instead
of always waiting for the next product improvements. If we keep
promising next years products what happens next year when we are working on
the following years upgrades and so on... When will it be good enough to
sell.
Anyway as Dennis Miller would say " I could
be wrong"
Thanks
Juan
If you have any thoughts about this now is the time to let me know.
Tab