EDA Competition Award: 1st prize of the IEEE CEDA for young Ilmenau scientists

Julian Kuners, student of Computer and Systems Engineering at the TU Ilmenau and student assistant at the IMMS, presented the paper "Trash or Treasure? Machine-learning based PCB layout anomaly detection with AnoPCB" to the jury consisting of representatives from Cadence Design Systems GmbH Germany, Dialog Semiconductor Germany, Infineon Technologies Germany, Gebze Technical University Turkey and the University of Reutlingen. The software project was supervised at TU Ilmenau by Dr. Marco Seeland and Professor Patrick Mäder at the Data-Intensive Systems and Visualization Group of the Department of Computer Science and Automation and at IMMS. The work builds on solutions developed in the Thuringian research group "IntelligEnt - Artificial Intelligence and Machine Learning for the Design and Verification of Complex Systems" for the layout of microelectronic chips. "When laying out analog/mixed-signal circuits, one designs the blueprint for the chip manufacturer. However, formally correct layouts can contain inconsistencies, such as substrate coupling and mismatch," explains Georg Gläser of IMMS, a specialist in integrating AI methods into design automation and head of the research group.

Design experience of engineers plays a major role especially in the geometric design of circuits and these last steps on the way to manufacturing require knowledge about which lines carry particularly sensitive or highly interfering signals and how they have to be handled, Gläser adds. "We have therefore developed an AI-based anomaly detection method in the research group that can detect non-proven and potentially faulty locations in layouts." The solutions for flexible data representation are important here, he said, because they can be used to process layout data for both chips and printed circuit boards - and the latter is what the award-winning contribution is all about.

Solution that can also be used in practical applications

"Julian Kuners, Henning Franke and Paul Kucera then further developed the software project as student employees at the IMMS. They put the finishing touches on our learning anomaly detection method as a plug-in for the free PCB design tool KiCad. This means that our approaches can be applied much more widely," is Gläser's assessment. The plugin allows KiCad signals to be categorized and passed to the training or evaluation process. The system was designed in such a way that the design data is prepared for the process at the user and then transmitted to a central server for processing. Thus, on the one hand, a possibly necessary graphics processor is only required in the server and, on the other hand, the designs of several users can be combined. The jury evaluated the candidates' solutions on the basis of complexity, degree of automation, designer interface, applicability, degree of integration with available design tools and robustness, among other factors: "The presented tool convinced the jury by the complexity of the posed problem, which in our estimation was solved well. The tool is user-friendly and we see it not only as an academic solution, but also as a solution that can be used in practical applications by PCB designers. The tool has considerable potential and we are interested to see how it progresses," said jury member Anton Klotz of Cadence Design Systems GmbH.

Award underscores high level of education at TU Ilmenau

Supervisor Marco Seeland from the Data-Intensive Systems and Visualization Group at TU Ilmenau is also enthusiastic: "The project impressively shows how the AI methods we have developed can be integrated into practical applications. The fact that the award-winning contribution originated from a project of our students simultaneously underlines the high level of education at TU Ilmenau and the enormous practical relevance of the skills taught."

The prize has also motivated Julian Kuners and his colleagues to continue working on the project: "For the first training, we used open-source designs such as Crazyflie and HackRF and then incorporated error locations there. With our anomaly detection plugin, we were able to locate these places quickly and correctly," explains Julian Kuners. "Of course, this spurs us on - and the prize anyway. We would like to use the occasion and call on developers to work with the plugin. The more training data there is, the more we can expand and improve it." The plugin will be made available for this purpose in the near future.

[Translate to English:]

[Translate to English:]