Electronic Voting Machines 1

[DonPE]

I'm just wondering what other engineers think of Electronic Voting Machines. There is a big push to implement these in the US, but there has been much controversy. There are advantages and disadvantages, but I was hoping to get engineers' perspectives on these devices, since some of us designed them (hopefully the design wasn't outsourced). There are several systems being considered, but I believe the push is for direct recording electronic (DRE) systems, which function like ATMs, and may or may not have a paper printout.

Do you think these systems are worth the cost of replacing the old mechanical ones? (Connecticut is moving toward DRE's and claims that it is hard to find replacement parts for the mechanical units).

What is the lifetime of these systems? It is only used once or twice a year, but may be sitting dormant in a basement or attic for the remaining time.

Can the devices be reprogrammed easily either in operational mode during an actual voting process, or while sitting around in storage?

Can secrecy of the ballot be maintained? Is electronic evesdropping/phreaking possible from either the keypad or the LCD screen?

Is the network secure, or will we likely see incidence of hacking & deleting files?

It seems to me that we are moving too quickly to a product that has not been thorougly designed and tested, and this could be a problem for our Republic.

 

[debodine]

Every single possible voting system except for all the masses standing in the same room, and everyone calling out their vote one at a time in front of everyone else, and everyone counting all the votes simultaneously and comparing the totals at the end, is open to manipulation by the vote counters.

Every system that adds a layer between the voters and the counters makes the tallying easier in a society of honest people, but conversely creates fewer and fewer people to "pay off" in a society of people who do not practice honesty to obtain a fraudulent result.

That having been said, you rightly raise serious questions about the reliability of an electronic system.

Considering the near genius level abilities of some of the top level hackers, the more we centralize on an integrated method of counting (think a network of voting machines statewide in any typical state), the easier it is for either outside hackers or inside manipulators to interfere with and modify the actual count. And the more centralized and compact the process, the less the number of points at which the interference must take place to achieve the fraudulent results.

Theoretically, if you can bribe the software creator to input a "back door" so that after delivery the creator can sneak into the system and set up the paid for result, then you only need interfere at one point in the system.

However, since I am neither a software guru or an electronic voting machine guru, I will wait to see what others who have more hands on experience have to say about the technical aspects of your questions. I have no actual experience from which to determine how likely it is that someone's tampering with software can be detected by others.

 

[itsmoked]

I think they are ridiculous. What a waste of money. Especially when society is heading towards mail-in ballots!

Take the pencil and fill in the little circle.... Just like we've all been trained to do in years of public schooling. Then there are no punched holes to even argue about.

 

[leonar40]

I wish you could just go online and do it. Of course that opens up a whole new set of possible ways that the vote could be corrupted or compromised. But it seems like if they have a way to encrypt your credit card transactions when you order something online, then they ought to be able to design a secure website for voting.

 

[ScottI2R]

DonPE,
As for your question of security against electronic evesdropping /phreaking being a concern. YES. This procedure is called "Van Eck Phreaking." And, as I just found out, it IS possible to "Van Eck" a laptop display. I was aware of CRT Van Ecking but I wasnt aware of the ability to snoop laptops that way. There are shielding measures that can be implemented to protect against EM monitoring.
Do a google on "Van Eck Phreaking" and you may like to do one on "Tempest Security" as well. The latter, a name coined by the US govt. regarding EM security.

Itsmoked,
Does this mean that if you didn't like the candidates, you could just pencil in a neat design from several circles? (Like we did in school when we didn't have a clue as to the correct answer?)

Scott


In a hundred years, it isn't going to matter anyway.

 

[itsmoked]

hahah Ouch! That was stewpid of you! You should have done what I did.. Pick "C" and do all of the "C"s down the page as you would probably get about %20 correct instead of z e r o.

:]

 

[ScottI2R]

AH yes, back to the laws of probability I see. Would you like to join me in Vegas itsmoked?
Scott

In a hundred years, it isn't going to matter anyway.

 

[itsmoked]

I'll be down that way on Friday! (Death Valley)

 

[Myndex]

Electronic voting machines are okay - and only okay - when they create a verifiable paper trail.

However, as a civil libertarian, I think a more serious problem is not the voting machines, but our voting system itself (plurality).

I wrote a couple of articles on the subject that you might find interesting:

http://civilliberty.about.com/cs/votingelections/a/condorcet012404.htm

http://civilliberty.about.com/cs/votingelections/i/ElectCol030804.htm

Andy

Andy Somers
Myndex Technologies Inc.

 

[jraef]

I agree with debodine in that every system has risks, and every system we have now has already had flaws exposed. I'm not a hacker, but I know some and they are connected to a wider underworld of hackers. Anything that can be programmed can be hacked, eventually. However, security algorithms can be so dense as to make it such a pain-in-the-rear to do it that motivation becomes a problem. Casinos are a great example. They are extremely vulnerable to hacking, yet it rarely happens. Average people can't be hackers (I'm a case in point of that), you must be extremely smart and smart people need motivation. Money (greed) can motivate some smart people to perform relatively simple tasks, but as a rule if you are smart enough to be a hacker, money is already not a problem for you. Religious or political zealotry can obviously motivate people to even kill themselves, but I have never met a hacker with that kind of passion for things outside of themselves. In general people smart enough to crack the toughest security algorithms are usually only motivated by the shear challenge of it.

Cases in point are the recent hacks of high level Pentagon systems. They now know they were hacked, but it doesn't appear that anything was done or stolen. Why? Because the ones who could have gained from it are probably not smart enough to do it, and the ones smart enough to do it have no ulterior motives. On the other hand, we as a society have plenty of motivation to keep the voting system secure, so even if the odd hacker/zealot were to find motivation to try it, they are (hopefully) up against a host of anti-hackers with paychecks and rightious motivation to stop them.

IMHO, the key to the entire system is to maintain the political will to invest heavily in the cost of keeping the system secure. My only fear is that the system will start out that way, but like most government systems, someone will cut funding later on because of some other agenda, and the security will break down. It can work as long as everyone understands the real cost.

"Venditori de oleum-vipera non vigere excordis populi"

 

[umn]

Electronic voting machines (EVMs) are cost effective and very much practical in the present world. Here're the most evident advantages (I'll come to the disadvantages too!):

1. No invalid votes
2. Result available almost instantaneously - based on authentication
3. EVMs are light weight, portable and easily installable. Less transportation costs
4. Forget those paper ballots- Means saving in paper and printing costs
5. Error free. No need for recounting of votes and the associated overheads
6. Tamperproof. No one can modify your vote or make it invalid after you cast it
7. The polling data is retained evenwhen the power pack is removed
8. Total secrecy of voting data. Requires special authentication to track/access.
9. The machines are reusable. They can be stored, reset and reconfigured for the next election.
10. Perhaps more secure than paper ballots. The modern digital encryption and digital signature technology combined with tamper-proof electronic hardware (processors, memory chips etc.) make it very much resistant to security attacks. (Any feasible attack would require a lot of computing resources, money and technical skills).
11. Can be locked by the presiding officer in case of booth-capture, which prevents any further casting of votes.

Disadvantages:

1. Initial investment is significant. But in a nation where democracy prevails, the pay back is assured in a couple of terms.

2. Not practical if the number of candidates are huge. In such instances, paper ballot is better.

3. One cannot offer 100% security (same is the case with paper ballots too!). If one can invest in enough computing power, sophisticated equipments and time, it might be
possible to hack the EVMs.


In India, use of EVMs has become the norm. For the Feb'05 elections held to three state Assemblies, roughly 40000 EVMs were used in 32000 polling stations. During the 2004 General Elections, 1.1 million EVMs were used. For the last election that used paper ballots, 7700 metric tonnes of paper and 2.5 million ballot boxes had to be used.

[um.narayanan@gmail.com]

 

[SolderingGunslinger]

One of the problems with all electronic voting systems is that the voting equipment is expensive. To require many small local governments to purchase or lease the equipment required could be a fiscal nightmare.

The single best answer is to use scantron ballots.

Why?

1) Most school districts already have scantron readers.

2) Most of the population of the US has filled out a scantron test at least once in their lifetime.

3) No new equipment need be purchased. Special voting booths are not required. Even absentee ballots can be read by scantron.

4) In the event of a close election, when many jurisdictions require a recount, there is a paper trail and paper ballots which may be examined by a canvassing board.

I remain,
The Old Soldering Gunslinger

 

[BJC]

People want to many things NOW, election results included. I can wait a week.
I don't mean to get involved in the poltics of the thing but the CEO of Diebold ( and the head of the republican party in Ohio ) promised to deliver the state to Bush.
Wheather he did or not is being debated still by others.
As to wheather it could have been done or not I have no doubt.
I don't even trust a machine that gives me a paper record. The correlation between my reciept and the totals in the voting machine cannot be established. You would have to get everyone who voted on that particular machine and compair the reciepts to the totals ( never going to happen)
Where I live we vote by mail. All the original ballots are in the courthouse they can be recounted again and again if necessary.
There are somethings machines just shouldn't do. Voting is one of them. If I lived in a state that used machines I would vote absentee ballot ( and encourage everyone else to do the same).