[Date Prev][Date Next] [Chronological] [Thread] [Top]

RE: Early Voting using AVTS - Shelby County, TN

Well written Jeff.  Attached are my thoughts on the card reader from around January.


One really good point you raised here though is that you needed to reboot the system to get the card out of the reader.  Could you get the card out of the reader with a paperclip instead?  Every reboot of the system in the field makes me “uncomfortable”.  It’s not that there is a specific problem with rebooting, but if we are going to corrupt the system some day, I bet it is going to happen with a reboot, so I would like to see operating procedures keep that to a minimum.




-----Original Message-----
From: owner-support@gesn.com [mailto:owner-support@gesn.com] On Behalf Of Jeff Hintz
Thursday, May 09, 2002 7:11 AM
To: support@gesn.com
Subject: RE: Early Voting using AVTS - Shelby County, TN


Sorry, I meant the new R-6 AVTS units.



-----Original Message-----
From: owner-support@gesn.com [mailto:owner-support@gesn.com]On Behalf Of Jeff Hintz
Sent: Wednesday, May 08, 2002 8:48 PM
To: Support Team
Subject: Early Voting using AVTS - Shelby County, TN

Here is a general summary for the Shelby County Primary Election on May 7th,  2002.  Shelby County used 133 new R-4 AVTS units for Early Voting.  They started on April 17th in the Election Office with 7 units.  Then on April 26th, they added their other 18 sites, with 7 units each.  They finished Early Voting on May 2nd.


They had 9 units that had problems during the Early Voting Process.  There were 3 units that would not release the Voter Card after voting.  It looks like 1 of the Card Readers has been installed too tight because when you insert a card there is friction when inserting the card and then when it ejects, it simply stays in the card reader, you then have to pull on the card while rebooting  the system in order to get the card out.  There are 2 other units that have no friction problem, but it looks like the card reader needs the spring or whatever ejects the card to be tightened/strengthened, because the card reader does not eject the card far enough out of the reader, and then the card is stuck in the reader, then having to  pull on the card while rebooting the system.  There are 3 units that do a continuous boot, when starting the system.  Once I inserted a PCMCIA card the continuous boot stopped, however the poll workers did not remove the PCMCIA card and reinsert it, which might have resolved the problem.  There are 2 units that went completely blank, and will not boot up.  There is 1 unit that boots, but the screen is a rainbow of colors and cannot be seen.


Obviously all of these units will be shipped back to McKinney for repair and maintenance.  I just wanted to give everyone that will be using AVTS units in the future, a heads up of possible problems you may encounter.  Also, this represents about a 7% failure rate.  Is there anything that the manufacturing plant can do to improve these problems.  If we install 19,000 units in Georgia, then we are looking at having to repair over 1300 units after their first election.


Other than machine problems and the Table Top Card Reader/New Smart Card problems.  Everything else went fine.  The loading up of all 133 PCMCIA cards, accumulating results, and Central Count all worked without a hitch.


Jeff Hintz
Diebold Election Systems
405 N. 115th Street
Omaha, NE 68154
(402) 697-7171 office
(402) 697-7166 fax
(402) 871-3346 cell


--- Begin Message ---
Just so that you are in the loop on the history...  The original prototypes did not have the locking reader that is in the units today.  They had a simple slot like the desktop card readers and the Jornada.  Here was the theory.  The on-screen instructions and pollworkers would instruct voters to leave their card inserted into the ballot station.  If a voter were to remove their card during voting, the screen would change to instructions telling the voter to reinsert their card to continue voting.  Though there might be some possibility of voters inserting and removing their card (like at a gas station), seeing a screen that says "insert your card and do not remove it" a couple of times would not leave a lot of room for confusion.  The beauty of this design is that it is light, cheap, has no special power considerations, and has no moving parts.
Enter Frank Kaplan who insisted that the reader have a locking mechanism.  About the same time, Jeff Dean took over research and development.  Between the two of them they decided to change the design.  Jeff and Ian found the locking reader we are using today, and changes were made to the housing to accommodate it.
It is obvious to me that the current locking reader won't work.  I have told this to anyone who would listen for about a year now.  The mouth takes too much manual dexterity.  It is non-obvious to people used to bank machines that you have to push the card with force the last quarter inch.  The locking mechanism is too lightweight to be reliable in a hostile environment.  In short, I don't think it can be fixed with a torque screwdriver.  Beware of those who do, or be willing to send them into the field when the weather changes.
That all said, I think finding a solution moving forward is going to be very "challenging".  We are probably too far gone at this point to return to the non-locking concept.  Perception alone will probably prevent that, even if it is the best solution in my humble opinion.
The alternative is to use a proper bank-machine-esque reader that "sucks" the card into the machine and provides the required reliability.  Unfortunately this is going to be large, heavy, and expensive, in a machine that is already too large, heavy, and expensive.
All just my $0.02.

> -----Original Message-----
> From: Green, Pat [mailto:GreenP@diebold.com]
> Sent: Thursday, January 24, 2002 2:11 PM
> To: 'Tab Iredale'; 'Ken Clark'
> Subject: Card reader update
> The likely cause of the card reader problems in manufacturing (see this
> morning's emails) is mechanical mounting.  Any stress placed on the reader
> seems to inhibit the operation of the release mechanism.  This
> isn't proven,
> but it's the best guess as of right now.   ...Pat

--- End Message ---