While patents and trademarks and all sorts of laws are designed to protect intellectual property, there’s a long road between making laws and enforcing them. In particular, global intellectual property theft is the most difficult to enforce due to all the ways in which copycats of any kind can slip through the cracks, and national borders often house some of those cracks. This isn’t always because of flaws in the system; there simply aren’t enough inspectors or the task at hand is limited by time constraints or human physics.
How big is the problem? The Korean Customs Service says over 30,000 seizures took place in 2019 at its borders, all of which were in some kind of intellectual property violation. And that’s just a drop in the bucket. Globally, over a billion dollars are estimated to be lost to counterfeits and piracy each year, and that number has been growing along with the number of industrial designs.
Detecting design infringements is now easier than ever before thanks to AI and short-wave infrared imaging (SWIR) cameras.
Imagine scanning a shipment of handbags that ‘look’ very similar to one made by a luxury brand. Now, unless each inspector is a handbag aficionado, they will need to consult a database, compare each registered design, then contact the IP owner to confirm – a logistical nightmare by any proportion.
However, using AI, a database of bag designs, and SWIR cameras, customs officials can vastly increase the efficiency of their operations. Any product can be scanned and analyzed against a database, and inferences can be made to determine whether or not a physical inspection may be warranted. This is preferable to random spot checks or relying on detective work before a container lands at a shipping yard, where time is of the essence.
A California based company, Stratio, Inc., has been working on AI that references a database of designs, compares them to what it “sees,” and assigns a similarity rating. “The AI inspection system compares photos of objects to images within an intellectual property database. Using this technology, we can eliminate the almost impossible process of manually sifting through all the similar designs and condense the work to just a few simple clicks,” says Co-Founder and VP of Systems Engineering at Stratio, Inc., Youngsik Kim. “Best part, this is all done from our simple, intuitive and time-saving mobile app.”
The development behind this use of AI is the result of a collaboration between Stratio and the South Korean National IT Industry Promotion Agency (NIPA) on the “AI Convergence for Illegal Counterfeit Inspection System” project. This was developed to help the Korean government identify design infringement cases at border checkpoints. With the support of NIPA, Stratio, Inc. began the AI developments two years ago. This included access to the country’s database of IP designs, as well as reviewing image data of previous infringements. Machine learning algorithms started by targeting the most commonly affected items at border checkpoints, including car parts, beauty products, IT products, and home appliances.
Much like how handheld devices have changed everything from buying groceries to buying homes, the convenience of a portable “infringement detector” can’t be underestimated from a resource perspective. As with all things AI, the goal is to fill in for humans in a way that scales intelligence and perception. The AI used in the camera helps refine the image and assist humans in figuring out what to look at, and why. However, at the border, some “hidden features” may be blocked from the AI. Thus, the AI is only part of the story.
That’s why in addition to the AI technology, Stratio Inc. has also been developing the world’s first smartphone-compatible SWIR camera, BeyonSense®. This will allow customs inspectors to look closer into a product without tearing it apart. The SWIR camera, BeyonSense®, is the first to use a proprietary germanium-based short-wave sensor for imaging in a more affordable, consumer-friendly package.
The technology has uses far beyond border controls. Intelligent scanning and detection has been in the news the last couple of years, like advancements in medical scanning that “see” cancer better than humans. It’s likely that Stratio’s core tech could be used in many contexts where “scan and compare” requires efficiency and accuracy – something that AI excels at. Add a more affordable, handheld SWIR camera like BeyonSense® and end users are able to unlock even more applications that require them to “see” into different spectrums. Stratio plans to launch BeyonSense® in March 2022.