Navigation
Topics
Log In
Log Out
:
Special Search
New Today
New This Week
Advanced Search
Tree View
Your Account
Edit Profile
Register
Forgot Password
Tools
Help/Instructions
Policies
...
|
| Author |
Message |
    
Jerry Berkman
Voting Rights Forum Participant
Username: Jerry
Post Number: 85
Registered: 5-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Monday, November 5, 2007 - 4:04 pm: | 
|
Brian Rothenberger of Monterey wrote me about some "mis-aligned"
buttons on an Edge touchscreen at a demonstration by the Monterey
Registrar of Voters:
> When the voter sees the touchscreen with a ballot in front of them,
> there is a BACK button and a NEXT button at the bottom of the
> touchscreen on the left and right sides, respectively. The buttons are
> rectangular in shape, about 3/8 inch high and 1 inch long. When I was
> touching these buttons, I found that the rectangular area of the
> touchscreen that would interpret the touch of my finger as a command
> to go BACK or NEXT did not match the rectangular area of that button
> that the voter sees on the touchscreen.
>
> "visual" button seen by voter:
> ________ ________
> | BACK | | NEXT |
>
> And on same horizontal position of screen, touching makes command
> happen outside the visual button as "touch" button is shifted to the
> left about 1/4 inch for the BACK command, and is shifted to the right
> about 1/4 inch for the NEXT command. Conversely, touching the right
> end of the BACK visual button or the left end of the NEXT visual
> button caused no response from the machine.
This appears to be a miscalibration. Why is it so hard to calibrate
touchscreen? This question probably applies both to touchscreens
for DREs and touchscreens for Ballot Marking Devices (BMD) like the
AutoMARK.
As I understand it, these "touchscreens" are really monitors with
a touchscreen attached on to it. The touch screen and the monitor
aren't logically connected; they both communicate with the computer
in the DRE or BMD, but do not communicate with each other. The
computer has to be told which coordinates on the touchscreen match
which coordinates on the monitor. In Brian's example, it may be that
the computer thinks the monitor is bigger than it really is.
I think ATMs have far fewer "buttons" on their touch screens than voting
machines and they are further apart. Maybe this is just the ATM
I usually use, or maybe it is a consistent pattern.
What are the variables in touchscreens? I noticed an ACCURATE project
report on touchscreens at:
http://accurate-voting.org/2006/11/01/touchscreen-calibration-issues-with-voting -machines/
which concludes:
quote: In the long term, the only solution is for the voting system
designers to use larger on-screen buttons, leaving more
room for calibration error. Alternatively, some vendors
avoid the calibration problem entirely by using hardware
buttons instead of a touchscreen.
So how do things like iPhones work? Do they use the same technology
or a more precise technology?
Also, on one page for buying touchscreens, you are given the option
to look by Monitor Type (CRT or LCD), by Max Resolution, by Input
(DVI or Analog), by price, by size, etc. Which of these
are the most important in preventing calibration problems?
Is it just that voting machine monitor are cheap and therefor
work poorly, or using old technology, or ??? |
V. Kurt Bellman
Frequent Voting Rights Forum Participant
Username: Formerelecdir
Post Number: 1566
Registered: 4-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Tuesday, November 6, 2007 - 8:42 am: |
|
Jerry,
The few vendors who have gone to hardware buttons are using much better systems. Touchscreens are inherently a problem. |
Richard Carback
Voting Rights Forum Participant
Username: Carback1
Post Number: 16
Registered: 10-2007
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Tuesday, November 6, 2007 - 1:16 pm: |
|
There is some useful information about touchscreen calibration here (it's old though, 2002):
http://www.embedded.com/story/OEG20020529S0046
NIST also has information about it:
http://vote.nist.gov/threats/papers/touchscreencalib.pdf
One way to deal with the problem might be to have voters calibrate the screen before they begin voting, e.g. "Touch inside the three circles to begin voting". This is not helpful if there's something maliciously making it miscalibrated, but it would fix innocent errors -- it is not clear to me that a mechanical solution would fix malicious interface problems (outside of making them definitely detectable versus "well it was miscalibrated" excuses).
I am not sure how the iPhone does it (I will ask somebody tonight). However, the palm uses touch the dots as above. |
V. Kurt Bellman
Frequent Voting Rights Forum Participant
Username: Formerelecdir
Post Number: 1568
Registered: 4-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Tuesday, November 6, 2007 - 1:27 pm: |
|
Richard,
The iPhone is a whole new technology. It resembles "traditional" touchscreen not at all. |
Richard Carback
Voting Rights Forum Participant
Username: Carback1
Post Number: 17
Registered: 10-2007
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Tuesday, November 6, 2007 - 8:13 pm: |
|
"The iPhone is a whole new technology. It resembles "traditional" touchscreen not at all."
I didn't know there was such a thing as a "traditional" touchscreen. To my understanding there are about a half-dozen to a dozen ways of doing it. In any case, I did make sure to mention, specifically, that I did not know at the time how the iPhone did its touch screen. So, I don't know what this comment was about -- I guess just a misunderstanding.
Anyway, it turns out that the iPhone uses a capacitive touch screen, which has a layer of material on the surface that stores an electrical charge. There are circuits in the corners of the screen, and when you touch it part of the charge goes through your finger. This decreases the charge on the screen and allows the computer to calculate the position touched on the screen. It seems to me that they could be built to be fairly accurate, although sometimes they need calibration (this is not really a great link but it talks about calibrating capacitive touch screens):
http://www.freepatentsonline.com/5283559.html
Wikipedia has more info on touchscreens in general, but there is not as much information as you'd like (e.g. how accurate are they?):
http://en.wikipedia.org/wiki/Touchscreen
I'm still waiting on more info about how the iPhone does its calibration. |
Brant Lamb
Frequent Voting Rights Forum Participant
Username: Brantl
Post Number: 1573
Registered: 1-2005
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Wednesday, November 7, 2007 - 4:49 am: |
|
In any setup where a proportional voltage is used to represent position, you are going to use a analog to digital convertor to render to the "brains" of the computer a binary value that represents the voltage sampled. There are two things that govern the accuracy in this system, the number of graduations that can be detected by the A to D converter and the fidelity/predictability and repeatability of the voltage made by touching a particular spot. So, if your touch screen physical materials are locked into your design, then the major factor left in accuracy is in how many times you sample to assure that your position sample is representative, and how finely your A to D converter can discriminate between voltage values.
So, accuracy is mostly a function of the consistancy of the performance of the touched material, and how many 'counts' the A to D convertor has in its range. |
V. Kurt Bellman
Frequent Voting Rights Forum Participant
Username: Formerelecdir
Post Number: 1569
Registered: 4-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Wednesday, November 7, 2007 - 9:41 am: |
|
Richard,
The "traditional" touchscreen I was referring to depends on physical pressure, a la a Palm screen, or a Kodak Picture Kiosk or similar. The fact that iPhone uses an electrical sensor, rather than pressure, is the difference I was referring to. Early systems (back in the late 1980's) used sensors in the bezel around the screen aimed across the screen's surface. |
Brant Lamb
Frequent Voting Rights Forum Participant
Username: Brantl
Post Number: 1575
Registered: 1-2005
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Wednesday, November 7, 2007 - 1:30 pm: |
|
A capacitive touch screen was designed by a friend of mine (who lived in Germany, before living here) and a partner in the 1980s. Most touchscreens (as far as I'd seen, I haven't kept up depended on two mildly conductive materials being brought into contact with each other. The use of 'electrical sensor' doesn't clear this up. |
|
|
