Genuinity Signatures: Designing Signatures for Verifying 3D Object Genuinity

Page Content

Research Areas: Prevention, Detection and Response

Principal Investigator: Mike Atallah

We are carrying out the research and evaluation of techniques for the manufacture of physical objects that support counterfeit recognition, tamper detection, and traitor identification. Counterfeit- recognition refers to the ability to distinguish an illegitimately manufactured copy of the physical object (even if it meets the manufacturing quality requirements). Traitor identification refers to the capacity to reveal the identity of the culprit user for an illegitimately copied physical object. We also adapt the counterfeit-recognition techniques to the related problem of tamper detection, which refers to the capacity to detect unauthorized modifications of the object. This is the first attempt at a science for marking objects whose copying is inherently imprecise, unlike digital objects that are replicable perfectly (with zero error). The approach we propose for achieving counterfeit-recognition consists of using a secret key (which is a large integer) to manufacture objects that carry a genuinity mark, such that the mark is automatically readable by a reader that does not have possession of the full key, and that operates from a single view of the object without requiring the object to have a particular orientation relative to the reader. The challenge is to design a genuinity mark that is not replicable by an adversary who does not have the secret key, even if the adversary has a golden model of the object and has manufacturing capability that is superior (i.e., has higher precision) to that of the legitimate manufacturer.   A re-instancing attack by such an adversary consists of using the golden model to produce the counterfeit copies (by making them close to the golden model).  A copying attack, on the other hand, consists of digitally acquiring the object from an available legitimate copy and then re-manufacturing it.   We seek solutions that thwart both types of attacks. Tamper-evidence will be achieved by using a separate mark that is appropriately fragile in the sense that damaging a legitimate instance of the object will destroy the mark --- this fragile mark is computed after the genuinity mark is determined, hence the genuinity mark must have enough resilience to withstand the changes necessitated by the subsequent modification that inserts the fragile mark (yet it cannot be so resilient as to persist in the face of a copying attack).  This project holds the promise of helping overcome counterfeiting of physical parts, a problem that has been called "the crime of the century" by a recent manufacturing industry report and whose yearly cost is rapidly escalating (its cost to the automotive industry alone was $12 billion in 1997).

Personnel

Other PIs: Daniel Aliaga (Purdue University)

Keywords: counterfeit, genuinity, mark, tamper detection