In the past year, we’ve seen the broader adoption of countless technologies spurred on by the pandemic. Many who were resistant to online shopping prior to 2020 have learned to embrace it. Video conferencing, once a last resort for collaboration, has become a mainstay. In my previous column, “The Dark Side of the Chip Shortage: Counterfeits,” I addressed one of unanticipated outcome of the crisis: the shortage of electronic components and predictable wave of counterfeit components likely to flood the market. Combating that tsunami of fakes may also accelerate the adoption of advanced techniques for detecting counterfeit components.
Electronics manufacturers, as well as component distributors, have developed a suite of tests, a combination of destructive and non-destructive techniques for establishing the authenticity of the parts they use. For those applications that require 100% inspection, no technology has exceeded X-ray inspection for its ability to validate parts. While as an industry we’ve successfully progressed from manual inspection to sophisticated automated parts inspection, it may be time to make that next leap. With the tremendous demand for a wide variety of components, and incentives high for counterfeiters, the time has come to supercharge component inspection by applying artificial intelligence (AI).
X-ray inspection is currently used to evaluate electronic components for many of the telltale signs of counterfeit parts. Inconsistent die size, inconsistent lead frame, missing die, and inconsistent die attach voiding are all internal features that can be indicative of suspect parts. Indications that a part has been reworked (possibly recycled) or otherwise mishandled include substantial die attach voiding, BGA voiding (possibly a byproduct of reballing), and bent leads. While these techniques have been effective, the current environment may demand that we up our game.
Figure 1: X-ray inspection easily reveals the counterfeit part.
For those who are disappointed with what we’ve seen from AI to date, it may have less to do with its capabilities than with the hype surrounding it. AI is at its best when applied to tasks that exhibit certain traits. These are typically narrow in their application, but substantial in terms of the data utilized, particularly in the learning process. To put that into context, the Tesla fleet delivers driving data from every connected vehicle in use. This data is more than one million hours of driving, which is greater than a single person would experience in a lifetime. Another example is the creation of a supercomputer to beat the world masters at the board game “Go,” which most people agree is the most difficult to play. The AI system could assimilate thousands of games and learn from them, resulting in an amazing level of “narrow” intelligence, that could outplay even the most experienced human. This is entirely different to human intelligence, which is much broader and complete, something contrary to sci-fi movies, is a long way away.
Narrow, Broad, and General Intelligence
So, narrow intelligence is when large sets of data are used to teach an AI system or a computer a simple task. The computer will seem extremely intelligent in its specialist subject, but dumb in just about everything else. This can and has produced amazing results in tasks like early disease diagnosis or predicting trends in supply and demand.
Broad intelligence is less simple and requires the AI to perform a set of tasks, thereby learning broader skills, but still within set parameters. An example of a broader system might be Amazon’s Alexa, or Apple’s Siri, with thousands of skills within limited parameters.
General intelligence is much more “human” and currently completely out of reach for AI. This level of cognitive thinking is the stuff of future dystopias, run by computers who manage humanity with a rigorous set of objectives, resulting in decision that seem cold and heartless.
A digital X-ray image contains a vast amount of data, not all of which can be interpreted by the human eye. Such digital images, and the rich data they represent, can be applied to component validation and counterfeit interdiction by using them to create a “digital fingerprint.” Of course, component fingerprinting already exists in the form of taggants, and have contributed to counterfeit interdiction efforts. But these taggants, often unique chemical agents, require expensive equipment for scanning, as well as training and certification. While useful, taggants lack the ability to identify features such as excess die attach voiding, which can be indicative of an otherwise authentic part that has been reworked, and thus allow counterfeits to corrupt parts inventories.
By leveraging the narrow intelligence capabilities of AI, algorithms trained on matching components to their fingerprints can rapidly authenticate parts, and can do so with X-ray equipment that most electronics manufacturers and component distributors already utilize. The combination of AI enabled software with automated digital X-ray can provide fast, highly accurate, autonomous screening of all possible internal and external features currently used to confirm authenticity of components. It’s no sci-fi robot detective, but it would surely strike fear in the heart of any counterfeiter.
Dr. Bill Cardoso is CEO of Creative Electron.