Objects have fingerprints. Sort of. Many manufactured objects contain unique characteristics that come directly from the manufacturing process - and a Swiss firm, AlpVision (founded in 2001, ten employees) wants now to exploit these "prints" to fight counterfeiting, using a sophisticated digital imaging software and analysis algorithms matched with Internet or wireless connectivity.
I've been e-mailing with the company's co-founder Roland Meylan, and here is how they explain it: "a machining or moulding leaves unique “prints” related to the tooling and the raw material used (plastics, metals, leather). There are always unique, microscopic differences on each and every part of a product. And these differences can be used to securely identify the object". This being Switzerland, the obvious application is counterfeited watches - a thriving market, particularly in Asia and online. Fake luxury watches are increasingly sophisticated, hence more difficult to spot. Besides advising customers to buy only through official resellers and stay away from Internet offerings, some watchmakers have tried to add holograms or seals to the packaging - but forged holograms didn't take long to hit the market. Suggestions to add RFID chips to the watches didn't catch on.
AlpVision's founders discovered that watches, even those machine-made, are unique in that they have microscopic differences. These for example are pictures of the back of two (apparently identical) steel watches, produced by the same machine. Notice the contour line of the logo "J":
These are very tiny, inherent manufacturing anomalies, invisible to the eye, that AlpVision wants to exploit digitally to identify watches. "Manufacturers keep a secure database of images of every genuine watch they produce; a buyer can scan a watch, or take a picture with his digital camera or cell phone, and send it in over the Internet to be compared with those in the database", explains Meylan. "Digital prints of previously identified counterfeits would also be kept, and the comparison is a matter of seconds". Reliability of the comparison is "excellent" with average office scanners, says Meylan, and the constant increase in camphones resolution goes in AlpVision's direction.
AlpVision says that the same system can be applied to the authentication of other products - for example, plastic casino chips and tokens. Here, the "fingerprint" would be unique for each series of chips produced from a same mould.