Researchers at the University of Florida have invented a new way to detect recycled and counterfeit electronic parts that makes supply chains more secure, protects consumer safety and runs at almost zero cost.

Their patent application, published by the U.S. Patent and Trademark Office on March 18, describes a technique for discovering when an electronic chip is recycled and seeing whether the recycled part is counterfeited. The team describes their method as a "universal testing technique" that can be used on a wide range of electronic components from different suppliers while costing virtually nothing to run. The candidate invention tests the power supply of a regulator in two common electronic parts: integrated circuits and system-on-chips, which combine different elements, such as video processing and memory, on the same physical chip.

Even at the Pentagon — the top military headquarters in the U.S. — about 15% of all purchased parts are counterfeit. And "it's not just the original manufacturers" that are affected, co-inventor Domenic Forte said — it's everyone. Though it is difficult to say exactly how much these fake and tampered-with products cost the public, estimates from the Electronic Resellers Association International suggest that "consumer and industrial businesses lose approximately $250 billion each year." If this number were translated into employment, it would represent over 36,000 lost American jobs.

Currently, all the ways to detect and prevent counterfeit electronics have a major flaw: They cannot work on all types of chips from different vendors. Some electrical tests focus on the performance of the parts, but they are not automated and often need data that is unavailable. Others use secure computer functions called hardware security primitives to test the behavior of a chip. The programs analyze the chips' authenticity to see if they are cloned or recycled. This approach, however, mainly works on new chips. A third method is physical inspection, which requires expensive imaging tests and expert technicians. Forte, also an associate professor at the University of Florida, told The Academic Times that he "wanted to do something different."

After a brainstorming session, Forte's team realized that they could use a certain type of regulator that is almost universal to see whether a chip is counterfeit. The team's invention tests the aging of these low-dropout regulators, or LDO regulators, and can work on electronic parts across the industry. LDO regulators modulate noise and ripples from the incoming power supply to produce a stable, fixed voltage. One way to measure LDO regulators is to see how capable they are of rejecting ripples in the input power supply. This performance metric provides numerical data that can help to "determine if the component is authentic or counterfeit," Forte said.

Forte and his colleagues used machine learning to "classify chips as recycled or new" after observing how much the power supply of LDO regulators was degraded. "While there are many types of counterfeit electronic components, recycled ones are reportedly the most common," the researchers reported in their patent. Many recycled components are removed from old circuit boards and then resold on the market. The team's computer program provides a safeguard against questionable electronics by determining whether the parts are likely counterfeit. 

"A lot of manufacturing happens offshore now," explained Forte, who added that this was especially true during the past year, with the pandemic. Manufacturers "worry about our entire supply chain — not just electronics — being routed and sourced through Asia, and there's a movement to develop more fabrication facilities in the U.S. again for chips," he said. Having more control over the process won't solve all of the problems in hardware security, "but it can solve some of them." 

The team has two possible scenarios to make this technology available for both consumers and companies. As Forte explained, the current invention requires measurements to be taken from every new chip, so researchers can retain the LDO-related data they need. "When you encounter the chip in the field, you would take the measurement again, and if there's a large deviation, you know that it was probably used," Forte said.

Another approach combines the technology in this patent with a self-referential method in LDO regulators. "It's an idea that as a chip ages, you have a way of measuring how much it aged," Forte explained. The researchers could date the component by comparing separate circuits on the same chip. They place two measurement tools on a chip, one of which serves as the power source for the electronics, while the other is just a reference for the behavior. The team in Florida is currently working on this technology.

Forte and his colleagues previously studied how counterfeit electronics are a threat to a safe economy. Looking ahead, Forte plans to use artificial intelligence and blockchain to support hardware security. In 2020, he was awarded a $1 million grant from the Army Research Office to secure electronic hardware systems with technology. Forte continues to work together on new projects with co-inventors Fatemeh Ganji and Nima Maghari. "The technology wouldn't have been successful without these great collaborators," he said.

The application for the patent, "Detection of recycled integrated circuits and system-on-chips based on degradation of power supply rejection ratio," was filed Sept. 1, 2020, with the U.S. Patent and Trademark Office. It was published March 18, 2021, with the application number 17/008722. The earliest priority date was Sept. 17, 2019. The inventors of the pending patent are Sreeja Chowdhury, Fatemeh Ganji, Nima Maghari, and Domenic Forte, University of Florida. The assignee is the University of Florida Research Foundation.

Parola Analytics provided technical research for this story.