Siemens & P&G Expand Their AI-Based Visual Inspection System Globally
The AI vision system inspects products at full speed, reducing scrap rates significantly.
At a Glance
- Real-time inspection adapts to material variations without extensive reconfiguration.
- Deep-learning models achieve full accuracy while inspecting thousands of products per minute.
- Deployment enables five-to-ten times faster commissioning than traditional systems.
In the last year or two, AI has gone a long way in proving its bona fides in quality inspection. As an example, Siemens and Procter & Gamble (P&G) are expanding their use of AI-based quality inspection. The quality solution developed in collaboration by the two companies is expanding across P&G’s manufacturing operations worldwide.
The system inspects products in real time during production, delivers comprehensive inspection coverage at full line speed, and helps improve quality consistency. Depending on the product, scrap rates have been reduced by 10 to 20 percent.
High-speed vision
The Visual Inspection Cockpit (VIC) is used in high-speed consumer goods manufacturing, where products consist of delicate, textured materials that naturally shift, stretch, or wrinkle at high production speeds. Traditional vision systems often require extensive reconfiguration when materials, packaging designs, or production environments change. By using Industrial AI, VIC can adapt more effectively to these variations and maintain inspection performance across a broad range of products.
The solution combines P&G’s deep learning models with Siemens’ Industrial Edge computing platform, industrial PCs powered by Nvidia GPUs, and AI hardware and software. Siemens supports the industrial computing infrastructure, software scaling capabilities, and long-term operation of the system across production sites.
Accuracy across thousands of products
"Our Industrial AI and Industrial Edge capabilities deliver what high-speed production demands: full inspection accuracy for thousands of products per minute, scalable from a single line to a global footprint. This collaboration demonstrates what we mean when we say we’re making industrial AI real," said Rainer Brehm, COO for automation and CTO at Siemens Digital Industries.
"We engineered this solution to solve a myriad of industry challenges traditional vision systems couldn't touch accurate characterization of overlapping components, the subtlety of low-contrast defects, the complexity of highly decorated products and packaging, tight time coordination, real-time PLC integration with single product reject at high production rates with continuous inspection," said Paul Thomas, director of machine vision and applied AI at Procter & Gamble.
Analyzing images in real time
The Visual Inspection Cockpit (VIC) analyzes live camera images in real time and inspects every product passing through the production line. Unlike conventional rule-based vision systems, the AI-based approach can handle a wide range of product variations without the need for frequent reprogramming. The solution includes the Visual Inspection Engineering Tool, which enables plant engineers to configure, train and update inspection models directly, without requiring dedicated data science resources. This makes quality control more flexible and easier to maintain in complex production environments.
Inspection results are processed close to the production equipment on Siemens’ Industrial Edge and integrated directly into manufacturing operations. The system can automatically trigger actions such as alerts or the removal of defective products from the production line. Quality data can also be collected and analyzed over time, helping production teams identify trends and support continuous improvement efforts.
Standardized applications
VIC is part of Siemens' broader machine vision and industrial AI portfolio, which includes the Industrial AI Suite. These solutions are delivered as standardized applications on the Industrial Edge platform. Together, they provide a consistent framework for deploying AI-based solutions at scale, regardless of whether the AI model is a custom development or Siemens-owned.Because VIC is delivered as a reusable Industrial Edge application, new deployments can be commissioned five to ten times faster than traditional bespoke vision systems. With established infrastructure, DevOps processes, and integration patterns, P&G can replicate the solution across plants, products, and inspection scenarios with minimal overhead. The inspection data feeds into P&G's broader digital manufacturing ecosystem, providing real-time visibility into process stability and continuous improvement opportunities.
Read more about:
Supplier News
Facts Only
* Siemens and Procter & Gamble are expanding their use of AI-based quality inspection.
* The system inspects products in real time during production.
* The solution reduces scrap rates by 10 to 20 percent depending on the product.
* The Visual Inspection Cockpit (VIC) adapts to material variations without extensive reconfiguration.
* Deep-learning models achieve full accuracy while inspecting thousands of products per minute.
* The system combines P&G’s deep learning models with Siemens’ Industrial Edge computing platform and Nvidia GPUs.
* The solution inspects overlapping components, low-contrast defects, and highly decorated products.
* Inspection results are processed on Siemens’ Industrial Edge and integrated directly into manufacturing operations.
* New deployments can be commissioned five to ten times faster than traditional systems.
Executive Summary
Siemens and Procter & Gamble are expanding a collaborative AI-based quality inspection solution across P&G’s global manufacturing operations. This system inspects products in real time during production, aiming to improve quality consistency and reduce scrap rates by 10 to 20 percent. The Visual Inspection Cockpit (VIC) uses deep learning models combined with Siemens’ Industrial Edge computing platform and Nvidia GPUs. The system is designed to handle high-speed inspection of products with delicate, shifting materials that present challenges for traditional vision systems.
The solution integrates P&G’s deep learning models with Siemens’ industrial infrastructure to provide full inspection accuracy for thousands of products per minute. It addresses complex inspection challenges such as characterizing overlapping components and low-contrast defects. The system allows plant engineers to configure and update inspection models directly via the Visual Inspection Engineering Tool, integrating results with manufacturing operations to trigger actions like alerts or product removal. Furthermore, deploying this solution via the Industrial Edge platform enables significantly faster commissioning compared to traditional systems, allowing for rapid replication across multiple sites.
Full Take
The narrative positions the AI solution as a convergence of high-speed processing, deep learning accuracy, and industrial infrastructure (Industrial Edge) to solve complex, dynamic quality control challenges inherent in consumer goods manufacturing. The emphasis on speed—inspecting thousands of products per minute—and adaptability addresses the core friction points of traditional vision systems, which require extensive manual reconfiguration for material or packaging changes. This frames the technology not just as an accuracy improvement but as a systemic shift toward adaptable, self-optimizing industrial processes.
The pattern observed is the leveraging of established industrial infrastructure to deliver advanced AI capabilities, suggesting that the path to broader AI adoption in heavy industry is through integration with existing operational hardware rather than entirely new bespoke systems. The claim of faster commissioning suggests an institutional advantage; this speed is achieved by standardizing deployment via reusable applications on the Industrial Edge platform, implying that organizational and DevOps maturity becomes as important as the technical algorithm itself for scaling complex industrial AI.
The implications point toward a potential bifurcation: organizations either adopt highly integrated, infrastructure-dependent solutions to gain efficiency rapidly, or they remain siloed with specialized algorithms. The focus on standardized applications suggests a pull towards ecosystem lock-in where flexibility is achieved through adherence to a platform framework rather than pure algorithmic novelty. What assumptions underpin the belief that integrating deep learning with edge computing inherently solves industrial complexity, and what are the long-term maintenance costs associated with this infrastructure dependency?
Sentinel — Human
The text reads like a professionally written press release or industry news piece, skillfully weaving together technical specifications with collaborative achievements. The evidence points toward human authorship focused on communicating joint technical advancements.