The continuing battle against counterfeit and inauthentic goods has a new weapon: the IBM Crypto Anchor Verifier.
Counterfeits siphon hoards of money out of the global economy, last year totaling $1.2 trillion in value with $98 billion of that coming from knockoff apparel, textiles and footwear, fake handbags and watches, and questionable cosmetics.
A March report released by the Organization for Economic Cooperation and Development found that proliferation of free-trade zones (FTZ) worldwide is directly correlated with the growth of counterfeiting. The more FTZs a country has, and the larger the zones, the larger the percentage of counterfeits coming out of that country, the report said.
“This is clear evidence that free trade zones are being used by criminals to traffic fake goods,” Marcos Bonturi, OECD public governance director, noted.
Not only do fakes and knockoffs erode brand profits, but they also can pose health risks and even harm the environment. Ten percent of all pharmaceutical drugs are counterfeits, according to World Health Organization estimates, and in sections of Asia, Africa and Latin America that figure rises to as much as 30 percent. Much of the extra-virgin olive oil (EVOO) traded worldwide fails to live up to “extra virgin” standards; in fact, EVOO is one of the most commonly faked foods out there.
With so many industries combating counterfeits, new tools are needed to stay one step of bad actors who evolve their tactics in hopes of evading detection. But with so many links in today’s supply chain, any useful new testing tool must be portable to each point along the way—and simple enough for any employee to use.
IBM Research came up with the Crypto Anchor Verifier, though its name belies its elegant application, to help the Gemological Institute of America (GIA) quickly authenticate its diamonds. At first, the GIA was excited about using blockchain technology to store digital records about the diamonds but then considered the next challenge: how to assure stakeholders that a particular blockchain record was tied to a particular diamond.
“I said, ‘we could build you an optics analytics mobile device that could uniquely identify physical product,’ and that’s how we invented the verifier,” IBM fellow Donna Dillenberger said. “That was the motivation—to tie the physical to the digital.”
On the back end, the verifier brings several technologies into play, including artificial intelligence (AI) and machine learning techniques such as video analytics and neural networks to power the software that calculates the optical characteristics of each sample. Coupled with a small optical attachment device that easily snaps onto smartphones like a paper clip and uses the cell phone camera to provide a “very microscopic resolution,” Dillenberger said, the verifier provides a complete authentication solution for myriad materials, calculating qualities such as color, hue saturation, motion and viscosity. With an item positioned beneath the verifier, the AI software calculates the profile, which users can then compare to a reference sample to confirm authenticity. IBM Research leveraged proprietary software and hardware optics to create the verifier, which functions even in low-light settings and captures an optical profile in subseconds.
“There’s been a lot of research behind what looks like a simple technology,” she added, noting that IBM Research spent years developing the AI model with teams spread across cities including Austin, Texas, New York and Zurich.
The verifier doesn’t automatically know the optical profiles of every object in the world, so it’s fed reference samples prior to scanning a new object. That means someone who wants to authenticate a passport, for example, must load a known sample into the software prior to scanning the document in question so that AI can learn the properties of a confirmed, genuine passport.
Though the GIA is the first mover with the verifier technology, Dillenberger said the device and software could give apparel professionals a tool to confirm authentic materials versus low-quality blends and fakes—especially for designers who pride themselves on using only the best textiles. While genuine premium wool and a synthetic blend might appear identical to the naked eye, the optical characteristics yielded by the verifier will indicate which one really is the real deal—and could aid apparel companies in ferreting out the weak links in their supply chain, holding suppliers accountable as well. A brand could upload a verified optical profile of a given item to the blockchain at its point of origin in the supply chain. If a later scan further down the chain reveals a different optical profile, that’s a valuable, hard-to-dispute clue that the item was switched or substituted somewhere along the way.
In addition, Dillenberger said users could include miniaturized QR codes on products labels—invisible to the human eye but readable with the verifier. These tiny codes could further serve as a means of distinguishing real products from knockoffs.
Although it was conceived to pair well with smartphones, the verifier also can work with other kinds of imaging devices that some businesses might already have in place. Users don’t have to store the generated optical profiles on the blockchain, either, if transparency isn’t their use case; if preferred, they can simply store the optical data locally on the smartphone itself, in a regular database, in the cloud, or on a laptop—wherever they choose.
According to Dillenberger, developments like the Crypto Anchor Verifier show how blockchain technology is encouraging innovation and new ways of thinking about old—and hard-to-solve—problems like counterfeiting.