Manufacturing has always been a balancing act between speed, quality, and innovation. For manufacturers in Consumer Packaged Goods, that balancing act is even more complex as product portfolios are expanding, packaging formats are evolving faster, and production lines are expected to accommodate more variation than ever before. At the same time, consumers continue to expect flawless products every time they make a purchase.
Proctor & Gamble (P&G) wants to ensure the best quality for its customers. Across thousands of products and production lines worldwide, even small changes in materials, packaging, graphics, textures, or production conditions can affect quality outcomes. Traditional inspection technologies were designed for an earlier world. Modern manufacturing demands something more adaptable.
That challenge helps explain why AI-based quality inspection is emerging as one of the most important applications of industrial AI. The ability to inspect every product at full production speed, learn from variation rather than be disrupted by it, and scale those capabilities globally gives manufacturers a new way to compete. Siemens' Visual Inspection Cockpit (VIC), developed in collaboration with P&G, demonstrates how AI, Industrial Edge computing, and machine vision can transform quality assurance from a production requirement into a strategic business advantage.
Powering quality inspection to address complex, faster manufacturing processes
P&G's fast production lines, with their high variability, are exactly the environment where the inspection process is demanding and requires a specific, smart solution. Production lines must accommodate high throughput, constant product innovation, evolving packaging designs, and materials that naturally stretch, shift, wrinkle, or vary during manufacturing.
Historically, rules-based machine vision systems have been used to detect quality issues. While effective in stable environments, these systems often require extensive retuning whenever products, materials, lighting conditions, or production settings change. As manufacturing variability increases, maintaining inspection performance can become increasingly resource intensive.
As Roland Joseph, senior technical director, Intelligent Controls and Automation at P&G, explained: "Traditional, rules-based vision systems simply struggle to meet the demands of the key product quality attribute inspections and the constant changes our innovation and manufacturing environments introduce. We need capabilities that are significantly more agile and robust."
What P&G needed was not simply a better inspection system. The company needed a fundamentally different approach, one capable of delivering continuous inspection precision at line speed while adapting to the realities of modern manufacturing.
Industrial-edge computing makes AI deployable at scale
Several years before partnering with Siemens, P&G had already identified the potential of deep learning to solve quality inspection challenges that conventional technologies struggled to address. The company successfully developed and deployed an internal solution across more than 150 production environments, proving the value of AI-powered inspection.
The next challenge was scale.
Deploying industrial AI globally requires more than developing accurate models. It requires standardized infrastructure, reliable computing power, operational resilience, lifecycle management, software deployment processes, and long-term support. These are the capabilities that transform a successful pilot into an enterprise-wide capability.
By partnering with Siemens, P&G combined its manufacturing expertise and deep-learning innovation with our Industrial Edge platform, industrial PCs powered by NVIDIA GPUs, and industrial AI infrastructure designed for deployment across multiple facilities and production environments.
"We engineered this solution,” Paul Thomas, director of machine vision and applied AI at P&G, said, “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."
The result is a solution that enables Industrial AI to operate where manufacturing happens: directly on the plant floor, in real time, and at speed.