Knocking Down the Bone Pile: Interposer Rework and Precision Solder Ball Attachment
Standard of Excellence: AI Makes It Possible, and Engineers Make It Happen
The Fearless Connector: Why I Chose the Hard Road
In early June, the EIPC Summer Conference in Vilnius, Lithuania, featured an impressive group of speakers ranging from PCB suppliers and technical experts to startups and robotics OEMs, the latter of which added a “cool factor” to the meeting’s ambiance with a fully operational, interactive humanoid.
After some early coverage of the event, I’m now following up with key takeaways from many of the technical presentations. I regret that I could not cover all speakers in detail, but this will give you a good taste of what you missed in Lithuania in June.
Honoring New Leadership
Rico Schlüter is indeed a charismatic and dynamic personality to lead EIPC into its next chapter. He assumed the role from Alun Morgan, who was honored at an evening dinner. Rico offers a different tone to his presidency and is, by all accounts, an industry “connector,” as Malcolm Gladwell would say. Rico is diligently working to bring all parts of the ecosystem together for more interaction and collaboration, the importance of which cannot be overemphasized at the technology and geopolitical inflection to which our industry is currently adjusting.
Following the morning coffee break on the first day, the conference focused on robotics and automation, starting with a thought-provoking presentation by Sultan Saidov of UK-based Alpha Machines. Drawing on more than two decades in AI, Saidov argued that the next wave of manufacturing innovation will be driven not by large language models alone, but by digital twin, where its roots stretch back to the Apollo missions of the 1960s. He noted that while early digital twins were used to simulate spacecraft, today's systems model everything from factory operations to near-Earth space, enabling organizations such as NASA and SpaceX to predict and prevent potential problems before they occur.
Saidov also highlighted Europe's continued leadership in foundational AI research, citing the pioneering work of Google DeepMind, yet he acknowledged that AI has had surprisingly little impact on PCB manufacturing, where every fabrication facility, process, and product is unique. That complexity has limited the effectiveness of current AI approaches. Looking ahead, he believes purpose-built AI models, coupled with digital twin technology, can gradually transform PCB manufacturing by capturing processes, equipment behavior, and production data, all creating a practical roadmap toward smarter, more adaptive automation.
Next up, Aurelijus Beleckis continued the discussion on robots. Read my interview with Aurelijus here.
Digital Twin and Plating
Reshmaa Selvakumar of Elsyca next explored how digital twins can improve one of PCB manufacturing's most critical processes: plating. She described a digital twin as a continuously evolving virtual model that grows alongside the manufacturing process, adapting as conditions and production requirements change. While today's statistical AI models excel at recognizing patterns from historical data, she explained that they struggle with novel or highly complex situations.
Mechanistic, physics-based models instead rely on fundamental equations to predict behavior, an approach she believes is better suited to the complexity of modern PCB fabrication. The ultimate objective is to combine both methods into a closed-loop digital twin capable of continuously learning and predicting outcomes. She demonstrated how digital twin technology can simulate plating scenarios and copper distribution, allowing manufacturers to predict through-hole plating, via filling, and potential void formation before production begins. The result is more uniform copper thickness, higher yields, improved reliability, and a shift from reactive problem-solving to predictive process control, even across multiple part numbers and board types.
Diving Into PCB’s AI Automation
In one of the next sessions, Dr. Günter Hoffmann, CEO of LexaTexer, an enterprise AI and predictive analytics company based in Berlin, Germany, discussed AI-driven audit automation. Richard Kujipers of Smart Robot Solutions, a leading European company in pick-and-place robotics, presented the company’s technology. Richard Nichols of GreenSource Engineering talked on the company’s groundbreaking work regarding some of the equipment offerings that help streamline manufacturing. Lukas Dinkelmann of Dyconex discussed sputtering and thin film solutions. Dennis Pusch of Schmoll Maschinen discussed their developments in a UV light source for PCB direct imaging (click here for full interview). Bas Le Grand at Nomenil, a startup company, continued the advanced manufacturing discussion with their work on process control for inkjet printing (click here for full interview).
Page 1 of 2
