OMRON Corporation (HQ: Shimogyo-ku, Kyoto; President and CEO: Junta Tsujinaga), an industrial automation company, announced in Japan an expansion of its integration of NVIDIA technologies by combining OMRON's Automated Optical Inspection (AOI) and 3D-CT X-ray Inspection (AXI) with NVIDIA Omniverse libraries. Specifically, this will make it possible to visualize substrate warpage caused by heat and to implement AI-driven "skillless" inspection, aiming to revolutionize the semiconductor inspection process.
Credit: OMRON
Achieving higher accuracy for skillless substrate inspection equipment.
Recently, as digital society has advanced, including the rise of AI factories, the innovation of various electronic devices, such as ultra-high-speed, high-capacity information processing, has accelerated. As a result, the mounting of printed circuit boards (PCBs), which support these devices, has become more complex, and advanced mounting technology and strict quality assurance are required. In manufacturing sites as well, the shortage of skilled technicians is becoming increasingly severe worldwide. Semiconductor manufacturers experience challenges related to substrate yields and quality control as a result of these shortages, making it urgent to shift to AI-driven quality control that can help scale individual expertise and know-how.
The technical integration of OMRON and NVIDIA AI and simulation technologies began in 2022. In this further collaboration, as part of an expanded, OT (operational technology) and IT (information technology) are linked to integrate manufacturing data that had previously been siloed, providing new insights (potential information) that lead to improved productivity, quality, and yields. Furthermore, OMRON will expand the application area of digital twins to PCB (printed circuit board/advanced semiconductors) inspection processes that require micron-level precision, promoting real-time optimization on the manufacturing floor. OMRON is engaged in the following three initiatives that leverage physical AI and simulation technologies.
■ Initiatives with NVIDIA
1. Visualizing board warpage through reproduction in digital space
“Board warpage,” which occurs when heat is applied during the manufacturing process, is one of the key challenges in electronic component manufacturing; while it can lead to solder joint defects and quality abnormalities in components, it is difficult to recognize with the naked eye. In this collaboration, we will reproduce and visualize "board warpage" with high precision in digital space. By integrating NVIDIA Omniverse libraries, OMRON can use physics-based simulation to reproduce board deformation and the force resisting deformation, enabling real-time visualization of subtle changes that were previously difficult to grasp. With these simulations, manufacturing sites can pre-adjust the time-based temperature changes (thermal profile) in the heating and cooling process, making it possible to predict and prevent critical solder joint defects before they occur.
- Investigating hidden causal relationships through the spatial overlay of AOI and AXI
OMRON uses digital twins to overlay the 3D surface data acquired by AOI (automated optical inspection) with the data on internal structures acquired by AXI (3D-CT X-ray inspection). This enables engineers to instantly identify the causal relationships behind phenomena occurring on the surface—such as pinpointing tiny internal bubbles that caused misalignment of surface components—and to make visible information that had previously been unseen, dramatically enhancing the essential value provided by inspection equipment.
- AI Agent Assisted Inspection
To address the global challenge of a shortage of skilled engineers and make inspections more AI-driven, OMRON will introduce visual inspection agents powered by NVIDIA Metropolis Blueprint for Video Search and Summarization (VSS) and NVIDIA Cosmos. VSS uses Cosmos Reason, an open vision language model (VLM), to analyze images, together with an LLM to understand historical data, enabling the agent to reason and propose next steps to users.
For example, when a user asks, "Why is this board warped?", the AI agent searches historical data for images of warped boards with similar characteristics, analyzes the root cause, and returns a detailed answer with insights to the user in an easy-to-understand summary.
"Leveraging OMRON’s know-how in automation rooted in a real on-site perspective, together with high-speed, high-precision control technology, we will realize highly accurate digital twins as part of the solution. By combining our inspection data with NVIDIA’s physical AI, customers will be able to visualize the factors affecting production, rapidly upskill novice operators, and maximize the ROI (return on investment) of production lines," said Motohiro Yamanishi, Company President of the Industrial Automation Company (IAB), OMRON Corporation.
OMRON has launched its medium-term roadmap, "SF 2nd Stage," from fiscal 2026 and has defined 13 focus businesses that will drive business growth. This time, Inspection Systems Business Division HQ is one of the core businesses that will drive innovation in manufacturing sites with NVIDIA physical AI. By combining its strengths across automation devices with inspection technologies, OMRON is delivering innovative inspection solutions for the semiconductor industry. Through this collaboration, OMRON aims to further strengthen its value creation for the semiconductor industry and contribute to solving customer issues to realize a more digitalized society.