Printed Circuit Boards Receive an AI Boost Through Fraunhofer’s Modular Platform

May 06, 2020 by Luke James

The AI platform from the Fraunhofer Institute for Applied Information Technology (FIT) is comprised of modules for artificial intelligence, machine learning, and deep learning.

A printed circuit board (PCB) is the platform upon which electronic components interact with one another. Today, they are found in a huge range of applications from the very simple to the increasingly complex. This has naturally led to more stringent requirements when it comes to design and quality assurance to ensure electromagnetic compatibility among other things. 

While this is great for consumers and users of electronic products, it is making the job of PCB design increasingly difficult for engineers who must make the best possible use of available space and position components closely together without risking failure. As such, when designing a PCB, great care must be taken, and the task often falls to highly experienced engineers. 

In a bid to help simplify and optimize the PCB design and testing processes, the Fraunhofer Institute has developed a modular AI platform that utilizes algorithms trained to undertake different tasks. 


PCB modular platform.

The modular platform with modules for machine learning, deep learning, and artificial intelligence. Image credited to Fraunhofer FIT


About the AI Modular Platform

“The modular design means we can harness several algorithms, which continually enhance their own performance. Data generated by ongoing automated inspection of components flows back to the algorithm. This then provides the basis for a process of self-learning by the artificial intelligence module,” explains Timo Brune, project manager at Fraunhofer FIT. “This permanent feedback enhances the database and optimizes the true negative rate. Early estimates from industry indicate this could reduce the use of production resources by around 20 percent.”

Once the algorithms have been trained, they can be used to design new PCBs from scratch. According to the team at FIT, this could end the “lengthy and costly procedure of trial and error” in PCB development where components are arranged and rearranged on the board until the optimal configuration is stumbled upon. The team also claims that the PCB application is just one example of many where a modular, self-enhancing algorithm can be used to enhance the electronic design process.