Hack resistant voting machine block diagram

Excuse the graphics, I whipped this up with the Windows Paint crapplet. The graphic is a bit simplified. The up counters will really be a set of bank switched counters. Each voting screen full of candidates would get a bank of counters. When the next screen is presented, there would be another bank of counters activated and the previous (and all other banks) rendered inaccessable.

The only communication between the COTS PC and the custom voting logic module is:

1. Counter bank setup information, multi-select (w/max permissible selects), or mutual exclusion mode, and an activation map for the buttons to activate.


2. A signal that the display information is on screen and the voter should be allowed to start pushing button(s).


3. Coming back to the COTS PC is a signal from the voting module that the current bank of selections has been committed

When a button is pushed, the corresponding LED will light indicating a selection has been made.

When the voter is done selecting for a particular page of votes, they hit the commit button. This tells the subsystem to strobe the counters corresponding to selections.

An additional (desirable) feature would be the ability to review/change/revisit votes. This feature can be added, by mapping the commit button to a set of bank switched flip-flops that would hold selection information. The addition of a "master commit" button would use the flip-flop banks to strobe the counters rather than the button state flip-flop's themselves. This would require an additional signal to the cots PC to "restart" the vote sequence for a voter.