Problems with Votomatic

M

Thread Starter

Mark Erdle

It seems that there are problems with the Votomatic HMI that is being used in certain election districts in the US. Some of these problems are ergonomic in nature, i.e. the operator punching an unintended hole and some are systems issues related to the pre-scored holes in the ballots, the 'hanging chad' problem. A National Bureau of Standards report stated in 1988: "It is generally not possible to exactly duplicate a count obtained on pre-scored punch cards."

Any thoughts on how to improve on this archaic HMI?

Mark Erdle
 
D

Dean Kindrai

It would appear that a great deal of the problems stemmed from the "archaic H" component of the "archaic HMI".

Besides, if a more reliable method of voting is instituted, a particular political party (I won't mention which) will have a harder time cheating.
 
Get Rid of It, Switch to Bar Codes -- we use them for everything else. Each bar code would would peel off, and would have the name of the candidate(s) on the label. This would be placed on the ballot, anywhere and be read. The bar codes would be scanned, and barring fraud in producing the bar codes, it would probably be the most accurate compared to all the other forms of voting.
 
J

John Coppini

No need for an improved vote counting machine! What we need are better voters and better media that report the news and issues accurately....democracy depends on an informed electorate and propaganda does not inform....

So, a better vote machine without the benefit of a better, informed voter would still suffer from....garbage in, garbage out.....
 
A

Allen Nelson

It's not an HMI problem, it's a Controls problem.

When I finish preparing my ballot, I put it in a box. Later, a machine (or human) reads it, and interprets my vote. I get no feedback on how my vote is tallied, or if I prepared it correctly.

What is needed is for me to run my ballot through a machine, let it tell my what it thinks I did. If it doesn't agree with what I thought I did, I make a correction. THEN I put it in the box. This is how we program our PLCs: with feedback. Why shouldn't voting be any different, regardless of what method (punchcard, barcode, optical scanner, etc) is used?

IMHO

Allen Nelson
 
P

Pradeep Chatterjee

In addition to the problem stated by Mark I feel another problem of great concern are rigging votes in some cases. In order to eliminate these
problems I feel we should seriously think of on-line voting systems.First of all identification of individuals won't be based on election cards or any other documents which can be forged. The voter should be identified with finger print detection system(on-line). Once he is identified he should vote on the screen of the machine (may be by clicking a mouse or using a "touch
screen"). The vote is submitted. Once the whole system is on-line within no time counting of ballots will be performed by computers and so there are no chances of human error.
These are some of my ideas and I am not sure if it's already available. If not then automation companies may think upon this (if feasible) and provide an error-free fair election.

Pradeep Chatterjee
M-30/7,
TELCO Colony,
Jamshedpur-831004 (India)
Email: [email protected]
Ph:91-0657-487549 (R)
 
K

Kirk S. Hegwood

Sorry, you're not going to fingerprint me without a huge fight. I'm not a criminal. Any automated equipment would have to use some other type of
identification.

Just playing schizophrenic devils advocate.


Kirk S. Hegwood
President
Signing for Hegwood Electric Service, Inc.
[email protected]
 
F

Francus, West (IndSys, GEFanuc, Albany)

check out Wired:

http://www.wired.com/news/politics/0,1283,40078,00.html

"... according to Ed Gerck, the CEO of the electronic voting firm Safevote
(www.safevote.com), Florida alone shouldn't be pilloried. He said that Palm Beach-type chaos occurs all the time with punch-card voting systems -- we just don't see it because the vote count usually isn't close enough to make a difference. "

"... Doug Lewis, director of the election-monitoring group Election Center (www.electioncenter.org), said that high-tech systems might create as many problems as they'd solve. "


West Francus
 
R

R A Peterson

This approach just puts the requirement for honesty on a different group of people - the people who program such systems. While on the whole programmers are probably more honest then politicians, this approach does nothing to
shield the system from the rampant corruption we see in many areas of the world, and in fact hides whatever audit trail the current paper based system provides making it far more difficult to prosecute election fraud.

The thing is that elections are political events and need to be shored up by the political process. A few years in jail for those caught messing with the system, and regular rotation of new people into the system is our only real hope. neither of which is likely to happen.
 
E
1. Practical solutions must deal with the voters AS THEY ARE.

2. Optical (blacken the designated circle) ballots seem to work very well and are an existing technology. They provide hard copy for manually recounting if required. Optical scanners could be improved to reach very close to 100% accuracy, but even the existing ones are much better than any punched ballot reader.

3. Punched card ballots should be illegal, period.

4. New technology in the voting booth could have many different approaches, all of which will work very well. I think certain requirements must be met:

A. A paper verifying copy of each person's vote
must be generated and placed in a conventional ballot box. These ballots must have a serial number that will match the serial number attached to every electronic vote (this would not compromise voter privacy since the system would not know who was voting).

B. The display presented to the voter must only allow one choice for each office which can be changed by the voter as many times as desired. This means that if the voter selects #1, the display indicates #1. If #2 is then selected, #1 goes off and #2 comes on, etc.

C. At the end of the selection process, the voter pushes the big VOTE button which sends the completed ballot to the tally computer AND prints out a confirmation of the vote. The confirming ballot goes into a ballot box.

D. If a voter is able to make a mistake with this
system, the election official can void a specific ballot by the serial number on its confirming paper ballot. This would not be needed often if the HMI is properly designed and implemented. A password system would identify and log the specific official doing any such correction.

This is not rocket science. It is simply a matter of our putting some attention on something that should have been replaced 30 years ago. If we could come up with a standard design for all states, it would become very affordable for all to implement.
 
D

Dave Ferguson

Not only that but as one person put it, We cannot even get certain people to follow an arrow, how do you expect us to get them to use a computer.........

Don't assume that because you use a computer every day, the rest of the world does also......

Our big companies cannot keep from being hacked into, do you think we would trust computer systems and corrupt politicians to keep the process honest. At least with a paper ballot people can understand a vote, put it in a
computer and everyone will think it has been tampered with, can't get anyone to use credit cards on web due to trust imagine a recount.......

IMHO

Dave Ferguson

Blandin Paper Company
UPM-Kymmene
DAVCO Automation
 
B
Votes are already being put into computers that have the potential of being tampered with. After all, that is the reason for the punch cards.This
discussion is centered around the input mechanism to make the mechanics of a vote, fair, obvious and above all simple. The techniques developed around lottery terminals fit this description. At least. I have not heard af any major problems with these systems.

Bob Pawley
250-493-6146
 
P

Pradeep Chatterjee

There were several opinions being put forward.
I feel first of all we should be clear that finger print detection is not a method to detect criminals only. It's a method to prove authentication. So with the scheme suggested finger print detection will help to identify an
individual of a particular constituency. Once it is authenticated the system will be enabled automatically to cast vote. Again it is locked automatically till other person is authenticated and the same person won't be allowed to vote again till next election is decleared.This will prevent rigging as political parties won't be in a position to capture forcefully the ballot
boxes or forcefully voting for other people. As far as security of the system is concerned it can be worked out in details to provide a possible
well secured system. If worked out properly it will surely provide a secured system than the ones existing today.

Regarding general people not being able to use a computer, shouldn't be a problem if we provide a touchscreen for casting vote. The screen will have
all the symbols of the parties and the vote caster will just touch on top of the symbol of his/her choice. In fact in India in some places electronic voting machines have been used where the caster presses a button by the side of the symbol of his/her choice. There wasn't much problem as such. Even initially if there's a bit problem people will definitely learn faster "to
touch screen" than the time they took to learn how to stamp when ballot paper stamping started.
As far as auditing is concerned the system can also be audited in cases of doubts but surely people conting ballots won't be able to do so. You will require software experts as auditors.In cases of extreme doubts provision can also be provided to take printouts of votes casted of a particular constituency and manual counting can be done to cross check.

In case of intentionally crashing a system there can always be a re-poll in that constituency (area)as it is even done today with the present systems in cases of several other problems. Moreover once a new system is started there will be several improvements to solve some other unforseen problems.

Pradeep Chatterjee
M-30/7,
TELCO Colony,
Jamshedpur-831004 (India)
Email: [email protected]
Ph:91-0657-487549 (R)
 
B
As an engineer, I have developed many Vision systems to look for holes or scratches and nonconformities of all size and shape. Most of these applications were very high speed and very accurate. Why cant someone develop a system that can at the very least Read the ballot corectly. Even with that many votes it wouldnt take this long to count all the ballots. Or recount and recount and recount..........

Bob Miller
 
A

Alex Pavloff

> There were several opinions being put forward.
> I feel first of all we should be clear that finger print detection is not
a
> method to detect criminals only. It's a method to prove authentication.

This is the major problem that many of the Americans (myself included) have with any sort of federal fingerprint system. The possibility for abuse is too great. It is the same reason why we don't have any sort of "United States ID card." There are many things that we don't want the federal government to know about us -- and fingerprints fall into this category.

I think the rest of your scheme, quite frankly, is overcomplicated.

Like any proper industrial project, let's write some specs here for what we want.

1) The voter must be easily to cast his votes for his desired candidate
2) There must be a physical output to allow for hand recounts and audits of votes
3) The system must be private and secure
4) The system must be able to quickly and accurately tally votes

(feel free to add any I missed)

Votomatic machines are in use because 1) they're paid for and 2) they pretty much work. I think quite a few of us have heard THAT refrain before, huh? Its become obvious now that indeed, they're not good enough.

As one of the other posters mentioned, a system where the voter, via a touchscreen HMI, picked his votes which were then punched or marked onto a
card of some sort (with perhaps, a quick non-certified count available over a network), would be the optimum solution to the inadequacies of our current system.

Going too far -- to the internet driven, paperless approach is, I feel, a step too far.

Alex Pavloff
Software Engineer
Eason Technology
 
B
Lottery terminals [have] most of the features needed. They are a proven technology.
An HMI which includes a feedback system so that the voter is assured of his vote, security ensuring only one vote per person and other normal security parameters are the only things that need to be included. This could be a first step to voting on all issues of import from the comfort of our homes and offices over the Internet.

Bob Pawley
250-493-6146
 
M

Michael Griffin

At 15:41 29/11/00 -0500, Francus, West wrote:
<clip>
>"... Doug Lewis, director of the election-monitoring group Election Center
>(www.electioncenter.org), said that high-tech systems might create as many
>problems as they'd solve. "
<clip>

I'm not really sure what problem the high tech election gear is supposed to solve. We just had a federal general election on Monday, using
the same system we've been using for over a 100 years. You mark an 'X' beside the candidate's name on a ballot paper and then stick the ballot in a box. After the polls close the election officials and the party scrutineers sit down and count the ballots.
An hour after the polls closed I was able to log on to the CBC web site and find out who won. Within four hours virtually all the ballots were counted across the whole country. I don't have any complaints about that sort of response time.

Before anyone starts designing a new election system, perhaps they should first be asking what the problem is they are trying to solve. Surely we've all seen enough examples from our own work to know that the highest tech solution isn't always the best one.
If you do insist that you want to put some computers in there somewhere, there is a company in Vancouver that has a touch screen voting system they are trying to sell...

By the way, if you want to rig an on-line voting system, I can think of a straight forward way to do this which doesn't require access to the
vote tabulating computer. Just crash, tie-up, or otherwise hinder the internet access points in areas where your opponent's supporters
predominate. Your opponent's supporters will give up, or else may not be able to log on until after the polls close. This may not work very well in
large densely populated cities, but many ridings are spread over large areas and have regional concentrations of support for different candidates and of course regional differences in internet access. Marginal ridings could be
swayed this way without anyone being able to prove what happened.


**********************
Michael Griffin
London, Ont. Canada
[email protected]
**********************
 
D
> Our big companies cannot keep from being hacked into, do you think we would
> trust computer systems and corrupt politicians to keep the process honest.
> At least with a paper ballot people can understand a vote, put it in a
> computer and everyone will think it has been tampered with, can't get anyone
> to use credit cards on web due to trust imagine a recount.......

You bring up interesting points. It's true that the computers that tally the results can be hacked, so there is a legitimate security risk with new voting technologies, and that is a good point to consider. More interesting though, is
that your response seems to imply that computers aren't involved in the punch card tally. I'm not too familiar with the process of counting the punch card vote, but I would bet that somewhere in the 19th century system used around the
country today, we are already accumulating vote totals on computers that can be hacked.

David L.
 
sometime around 1880 was the first instance of using punch cards and electromechanical counting machines in US elections, the precise date is a
bit foggy.
 
R
Any system for election should handle the demands of validation, tabulation, and audit. The prior suggestion of using bar codes (sorry, I do not have the footnote) seemed on the right track, but instead of using peel offs as suggested why not have the voting booth terminal just print it out. Here is how it would work.

The voter would receive a blank ballot, pre-printed with suitable counterfeit foiling, to take to the terminal. A touch screen allows
candidate selection and validates that the voter does not double select and queries any non-selection ("...are you sure you don't want either candidate for dog catcher?"), and finally prints out the ballot, complete with bar codes and the candidates selected in human language, say English. The voter drops the ballot in the box...next. At the end of the day the polling supervisor receives a code from the district
supervisor which allows each terminal to print a report of totals, which must be reconciled with the result of a machine read of the bar coded ballots. Voila...validation, tabulation, and audit.

The biggest barrier to such a system may be the capital expense. Anyone know of how to build such terminals for cheap??? Any refinements to
such a system??? You may not get rich but you might have the satisfaction of saving Democracy :)

Regards

Ron Sewell
Sigmatic Controls
Kelowna B.C.

Being the entire personnel of Sigmatic, I have a tough time maintaining that my opinions do not reflect those of my employer!
>>>
...I'm not really sure what problem the high tech election gear is
supposed to solve....

Michael Griffin
<<<<
 
Top