[E-voting] Battery of Nedap

Adrian Colley aecolley at spamcop.net
Fri Jun 10 16:09:22 IST 2005


On Thu, Jun 09, 2005 at 04:18:29PM +0200, Pierre Muller wrote:
> What happens if the power failure occurs exactly when the vote is recorded ?

Joe McCarthy mentioned this problem in his submission to the Commission
on Electronic Voting:
> PTB were not asked to test against the hardware tests of the following:
> 
>  8.2 Subject to 10.3, a cast vote must not be lost by a power failure,
>      the failing of one component, the effect of environmental conditions
>      detailed in paragraph 11.1, through normal use or through failures
>      in the operation of the voting machine.
>  8.3 The installed ballot module and its contents must be fully
>      maintained in case of a power failure, the effect of environmental
>      conditions as set out in paragraph 11.1, through normal use or
>      through failures in the operation of the voting machine.
>  8.4 The functions of the voting machine must be fully maintained in the
>      event of a power failure, or exposure to the environmental
>      conditions as detailed in 11.1.
> 10.2 In the event of a power failure between 500 msec and 2000 msec of
>      pressing the CAST VOTE(S) button, a vote stored in the Eeprom of
>      the machine shall not be lost, providing there is no fatal machine
>      failure on restoration of power.
> 10.3 In the event of a voting machine failure that affects the vote
>      storage process (between 500 msec and 2000 msec of pressing the
>      "CAST VOTE(S)" button), unless the failure is such as to cause a
>      complete shut down of the machine, an error message shall be
>      displayed e.g. error no. xxxx - vote not stored.
> 
> Note that the Department's reply stated "testing of these rules was
> undertaken separately by the UK-based Electoral Reform Services"
> indicating that ERS were to do these tests.  But ERS have no role
> whatsoever in testing the voting machine.  They have a very limited
> role in testing the counting functions of IES.  Therefore nobody has
> tested the voting machines for the above requirements.

Also in his submission, Joe quoted from the Zerflow report:
> Zerflow's third finding was:
> 
>     In the event of power loss, unless the control unit operator is
>     keeping a constant count of votes there will be no way of knowing if
>     a voter has actually cast a vote.
> 
> The response from the Department was "In the event of power loss, if a
> voter is in any doubt that their vote has not been cast successfully, a
> tally of the permit tickets and the number of votes cast on voting
> machines will provide a simple means of validating the vote count.
> The risk of this event occurring is minimal."
> 
> I wish to analyse the Department's response in detail.
> 
>     In the event of power loss, if a voter is in any doubt that their
>     vote has not been cast successfully,
> 
> It is not up to the voter to ascertain if his vote has been cast
> successfully.  It is the responsibility of the Minister, his Department
> and their suppliers to ensure this.
> 
>     a tally of the permit tickets and the number of votes cast on voting
>     machines
> 
> This phrase refers to "machines" plural.  Why is this?  The voter after
> all has only been using one machine.
> 
>     will provide a simple means of validating the vote count.
> 
> The implication of this phrase is that there is a direct correlation
> between permits to vote and votes recorded.  For three reasons this is
> likely to be untrue:
> 
> 1. The voting machine does not always record votes just after a power
>    failure depending on the timing between pressing the Cast Vote button
>    and the actual power failure.
> 2. The list of errors published in the Voting Machine Operators Manual
>    is shown below.  It lists 15 different message codes which might
>    arise each of which "means that the vote is stored but NOT counted".
> 3. A voter may walk away from the machine without voting thus no vote
>    will be counted for his permit ticket.
> 
> 			    Troubleshooting.
>     In the unlikely event that the voting machine does not function
>     according to the procedure described call the help desk number
>     provided.
> 
>     The display above the voters panel will, in most circumstances show a
>     message if a problem occurs.
> 
>     Make a note of this message so that you can tell the help desk. This
>     will help diagnosis of the problem and what action to take.
> 
>     If you see a message in the display which consists of one of the
>     following numbers it means that the vote has not been stored.
> 	8001 8002 5406 5410 5503 5504
> 
>     If you see a message in the display which consists of one of the
>     following numbers it means that the vote is stored but NOT counted.
> 	5104 5108 5112 5116 5118 5120 5122 5124
> 	5126 5128 5129 5131 5132 5133 5139
> 
>     The voting machine should be exchanged and the voter must be asked to
>     choose their preferences again on this new voting machine and press
>     the cast vote button.
> 
> The last instruction in the above extract from the Operators Manual
> means that the voter will be allowed to vote twice.  This is
> extraordinary.

It's the e-voting lottery!  Win a free extra vote -- it could be you!

 --Adrian.

-- 
GPG 0x43D3AD19 17D2 CA6E A18E 1177 A361  C14C 29DB BA4B 43D3 AD19
http://user-aecolley.jini.org/



More information about the E-voting mailing list