Hummm.... how about a bunch of phototransistors, carefully placed to 
pick up each segment in the 7-segment displays?  Certainly a hassle but 
a lot easier than doing OCR on a microcontroller.

Although if they don't encrypt their data, that might be even easier. 
Is there a chance they could 'shoot back' and update their systems to 
thwart you if you are able to sniff the RF?

I hope you succeed, companies that try to lock up systems and impede 
progress to help ensure their profit stream deserve to be bypassed and 
(hopefully) not get new customers as their lock-in philosophy becomes 
public knowledge.

Let us know how it works out...
J



Roger Weichert wrote:
> Hi Guys,
> 
> I was approached yesterday to offer a solution to a problem a 'herd testing' 
> company has with obtaining 'dairy cow identification numbers' from the RFID 
> system already built into the dairies.
> 
> They want a simple device they can clip to the front of a led display (7 or 
> 8 seven segment displays) that will transfer the number to their own 
> equipment.
> 
> They cant physically tap into the existing displays because they would have 
> to modify the display in every dairy they go to (which is impractical) and 
> the manufacturer of the equip already in the dairies will not provide info 
> on the protocols used to send the numbers to the display (despite 5 years of 
> negotiation)
> 
> Apparently the herd testers take their own equipment to each dairy to obtain 
> milk samples but it only stays there a couple of days. They return each 
> month.
> 
> That exists of their own RFID antenna and reader to identify each cow, then 
> a palm device and a printer to spit out a label to go on each 250ml sample 
> container.
> 
> They obtain two samples per cow from two milkings. The samples are then 
> taken back to the lab where they use a robotic sorter before being analysed.
> 
> Apparently they have investigated using a camera in a box, hung on the front 
> of the dairies display, and then using an OCR program to identify the 
> number.
> 
> I suspect the best way would be to 'crack' the communications feeding from 
> the reader to the display, though exactly how to do that is my dilemma.
> 
> Does anyone have any suggestions how to go about it, or any other ways of 
> capturing the ID numbers.
> 
> The two brands of RFID systems already in the dairies are made by DeLaval 
> and Westfalia.
> 
> Regards, Roger 
> 
-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist