Why Anti-Counterfeit Technology is Critical to Electronic Product Development
BY DR. EYAL WEISS, CTO AND FOUNDER OF CYBORD
The electronic goods we just can’t live without – our smartphones and computers, entertainment systems and home appliances, for example – all rely on the integral components that make up Printed Circuit Boards (PCBs). Unfortunately, the supply chain for these critical parts is still being disrupted in the aftermath of the pandemic, including those for bulk electronic components and their raw materials.
As a result, some manufacturers have passed up conventional supply channels, opting to buy components off the free market. It seems like an obvious solution, yet this practice predictably gave rise to supply streams of counterfeit goods. In fact, the shift away from standard suppliers has increased the risk of receiving out-of-date, mislabelled, poorly handled, or defective electronic components from the typical 0.2-1% to 5-10%.
While lab testing can be used to identify specific errors in such parts, it is ill-equipped to address the full breadth of component defects, either because they are scattered or appear under very specific conditions. Furthermore, testing is only conducted via sampling, and in most cases, acts to conceal counterfeiting are quite sophisticated, making it even more difficult to detect them from just a few components – not to mention, it can be prohibitively expensive and slow.
AI lends itself far better to detecting counterfeited electronic components. Without it, unsuspecting Original Equipment Manufacturers (OEMs) and Electronic Manufacturing Services (EMSs) risk civil and corporate lawsuits in the event of a malfunction-related injury, millions in lost revenue that could potentially tarnish years of hard-earned reputation, and crippling product recalls.
The Counterfeit Misfits
Counterfeit electronic components have infiltrated a wide range of industries, especially at times when supply chains are disrupted, and are therefore far more damaging than manufacturers would like to believe. They cause random, unpredictable statistical failures that are extremely difficult to detect through conventional acceptance tests. And even when imposters are identified, they are often dismissed as workmanship errors.
Simply put, the quality of raw materials is critical to the quality of the final product. This is why isolating compromised components before they in turn compromise their end products is so crucial.
The substandard materials used to create counterfeit circuit chips erode quickly, diminishing not only the reliability of the component, but the integrity of the product itself. In the case of vital equipment supplies within healthcare, automotive, or defense industries, for example, a malfunction could very well result in a life-threatening incident.
The conventional method to reduce the risk of unsafe components is by procuring them from authorized distributors – in short, relying on the supply chain of the Original Components Manufacturer(OCM). While materials and workmanship are distinct, there is no denying their co-dependency. Standard assembly practices, however, only monitor and analyze circuit board assembly processes for quality assurance, rather than the state of the components themselves.
AI-powered visual technology is a far more efficient and cost-effective solution than lab testing. Regardless of where components are sourced from, AI can rapidly analyze each and every component for authenticity without concern for human error.
SMT machines – the machines used to mount electrical components onto the surface of PCBs – already use visual alignment during assembly and produce onsite images throughout the assembly process. By leveraging the data already collected by these assembly machines, machine learning-based image processing can be used to root out counterfeit and defective components on the assembly line itself.
AI also provides brands with verifiable proof that all the materials used in the development of their product were meticulously checked at the time of placement. In the case of a failure in the field, data
analysis can potentially save companies from instituting product-wide recalls, thanks to the ability to vet and track every component on an individual basis.
With an expansive database of images, not only can AI systems be better trained to detect errors, anomalies, defects, and counterfeit materials.
For as long as there are supply chain issues, manufacturers will resort to sourcing electronic components from the free market. But without a better means of distinguishing compromised materials from legitimate ones, counterfeit electronic components will continue to permeate and jeopardize the electronic products and devices we rely on, posing risk to revenue and, more critically, public safety.
AI’s capacity to identify the subtlest changes in detail in real-time offers a reasonable and practical means by which to ascertain authenticity efficiently and cost-effectively.