Issue link: https://iconnect007.uberflip.com/i/1541670
42 SMT007 MAGAZINE I DECEMBER 2025 These answers suggest that custom- ers are seeing value in smarter and/or faster inspection machinery. In North America specifically, these trends seem to be driven by the need for capability and capacity associated with reshoring. We wanted to know what specific features are currently driving their sales. Phung said everyone is asking about AI technologies, and that Koh Young is in a unique position to leverage AI software in its systems to further automate processes and elimi- nate quite a bit of human interaction, which can also enable even more machine-to-machine communi- cations. "Koh Young also offers proprietary machine-to- machine AI solutions, which embed process knowl- edge directly into the software," he said. "This enables inspection systems to automatically adjust upstream process equipment, ensuring consistent quality and optimizing the process. We're devel- oping multiple AI and ML solutions designed to achieve smart factory objectives." Fieldhouse said that in some cases, "we're deal- ing with more complex PCB assemblies at Omron, products that require more or different inspection than previously needed. Many of these applications have smaller features and solder joint dimensions, so our customers require a more capable machine." With both Fieldhouse and Phung mentioning AI, along with speeding up inspection and improving capabilities specific to smaller component sizes, we continued asking questions about what's driv- ing sales in the different market sectors and how OEMs are changing what the EMS companies are shopping for. "The demand for X-ray inspection has been increasing over the years, largely due to OEMs requiring 100% inspection coverage and the increased use of BGAs," Field- house said. "When it comes to inspecting the complex compo- nents used in today's generative AI server PCBs, many EMS compa- nies are forced to upgrade their old 2D or 2.5D systems to a more capable 3D CT system. Omron's platform can pene- trate thick heatsinks or PCBs and iden- tify solder defects hidden under large or small BGAs." Fieldhouse tied together both the hardware and the software side of AI in the current inspection market. Artificial intelligence and machine learning technologies are being employed within the manufacturing equipment to build the boards used in AI compute centers. Koh Young has a reputation for implementing machine learning into its software, but how has this AI/ML capability opened business opportunities for its customers? Phung said his company has been proactive in adopting a variety of ML and deep learning tech- nologies, including generative AI. "Some of these solutions operate seamlessly in the background, providing enhanced functionality that may not be immediately visible to the end user," he said. "Others significantly improve equipment usability through automated programming and automated defect review, contributing to improved efficiency and potentially a smaller operating staff." Naturally, he added, the value of these AI/ML solutions depends on each customer's production environment. "High-mix manufacturers benefit from significantly reduced programming time through automated programming, while high-volume customers see substantial reductions in labor requirements and operator errors. Those utilizing our KPO tools enjoy real-time production moni- toring and automatic process adjustments driven by AI/ML engines. We view this as a genuine competitive advantage, enabling us to minimize defects and maintain supe- rior process quality." Fieldhouse mentioned 100% inspection coverage and BGA components. New packaging technologies being developed in semiconductors and compo- nents are diverging trends. On the one hand, cutting-edge passive components nowadays are essen- tially microscopic, while cutting-edge A l l e n P h u n g N i c k F i e l d h o u s e