When we think about the opportunity for edge computing with artificial intelligence and manufacturing, we see a huge opportunity to transform that business. When we think about it in terms of edge computing with artificial intelligence, we think of four examples. One being, machine vision for defect detection. We also look at it for worker safety. We also look at it for condition monitoring, or even predictive maintenance.
And we're not only transforming the outcomes of these businesses, but we're delivering a transformational path to how we deliver them. If you look at defect detection and some of those outcomes, we're able to take machine vision and transform how a factory delivers, not just defect detection for higher quality, but also faster throughput, less scrap, and, ultimately, lower opex. We're literally, in some cases, replacing the human worker with off-the-shelf IT capabilities to deliver on that industry 4.0 promise.
When factories rely on human eyes to perform precise inspections, the result can sometimes be less than optimal. In some factories, less than 20% of defects are actually detected. And that means expensive customer returns and wasted machine and worker hours.
So we asked ourselves a question. What if Intel's data analytics and high-performance computing technology could enable machines to evaluate product quality on the factory floor while the work is in progress? We turned our eyes towards creating a scalable solution with commercially-available hardware and Intel software. And we tested it in the fast-moving China market.
It all starts with a camera. A high-definition, off-the-shelf camera is used to inspect the part from all angles. The data streams to an industrial PC and it's powered by an Intel Core i3, i5, or i7 processor. And these are all compatible with the full suite of Intel-based accelerators with algorithms created using Intel's OpenVINO, computer vision, and deep learning combined. And we use that to evaluate whether the part passes inspection.
Other industrial machines in the factory can be connected to Intel-powered gateways that gather and translate data to standardized formats. Then they transfer that over the network to either the edge server or the cloud for processing. The edge server, powered by Intel Xeon, scalable processors, sorts and processes the data using Intel's industrial Edge Insights software, identifying which images and data streams need to be stored. The edge server acts as a scalable, on-premise data center in the factory, processing data from multiple machines, consolidating workloads at the edge, and sharing data with the cloud.
Factory workers can use ruggedized tablets to check the status of operations. Data is collected over time and is used to identify and address the root causes of defects. The results mean time and money saved on the factory floor. A proven industrial edge computing platform, designed to be scalable, ready for integration into any factory, and already implemented in China.
Camera technology is being deployed ubiquitously, giving computers the ability to see what is happening. When paired with recent innovations in AI technology, the computers are also able to use deep learning to understand, or infer, what they are seeing. The technology is transforming business outcomes across industries, making cities safer, modes of transportation and manufacturing facilities more efficient, driving increased sales and operational efficiencies for retailers, and providing life-saving opportunities through early detection in hospitals.
The possibilities of using computer vision to drive business transformation and outcomes are endless. Please reach out to your Intel account representative or visit us at intel.com/iot. We look forward to partnering with you.