• Electronic components counterfeit mitigation by 100% deep visual inspection

    cyborg

    Allocations are forcing manufacturers to purchase components in the free market. This is increasing the risk of counterfeit, out-of-date, mixed lots, badly handled, recycled, and defective components from the typical 0.5-2% when purchasing from trusted sources to 5-10% when purchasing in the free market. While this risk is conventionally mitigated by performing sample lab tests on components purchased in the free market according to SAE AS6081, sampling a small fraction of the components is simply not enough. Lab testing does not address out-of-date, mixed lots, badly handled, or defective components which may be randomly scattered within a package. In addition, the process is slow, inaccurate, and expensive.

    The packaging of components has subtle intrinsic features imprinted during the manufacturing process that serves as a “fingerprint” of the source of the components. Recycled components may undergo black-topping, remarking, and repacking that leave visible evidence on the component exterior. Cloned, overproduced, and defective components are packed using a different source, presenting a different package. Re-dating old components may be detected externally by the poor state of the leads. These subtle features may be employed to authenticate components.

    Cybord introduces a novel method that allows checking components visually and evaluates their source, quality, and trustworthiness. The images are processed on the fly by a series of artificial intelligence algorithms capable of authenticating the manufacturer, determining the conditions of the soldering leads, finding defects and evidence of unauthorized firmware programming.

    Packaging of counterfeit components today is designed to avoid detection by sampling. An example can be seen in the figure above, where production images taken by the pick-and-place machine vision system (ASM – Siplace SX) and extracted by Cybord SMT, showing that the beginning of the reel is authenticated to the documented manufacturer (Taiyo Yuden in this example) and the rest of the components on the reel are manufactured by an unknown source (counterfeit) as also exemplified by the example images and the measured dimensions. The spiral on the left side represents the position on the reel where each color stands for a different detected group. The outer shell of the reel is detected as 0603 MLCC made by Taiyo Yuden and the internal part is of an unknown source.

    Cybord also offers an incoming material inspection solution called Cybord Lab, authenticating all the components in the reel of any type and package in under 15 minutes before they enter production. The method authenticates and detects mixed lots and date-codes in a reel by deep marking analysis.

    The shortage in components may be the turning point in production history where component inspection analytics I4.0 advanced technology replaces manual, out-of-date methods and thereby improves productivity, reliability, and security of electronic products.

    Eyal Weiss, Ph.D.

    CTO and Founder at Cybord I Quality and reliability of electronic components by Big-data and AI.
  • Subscribe today to the industry leading magazine and newsletter

    Keep up-to-date with information about worldwide trends straight to your inbox or through our printed magazine.