Skip to main content

ARTIFICIAL INTELLIGENCE

Catch Defects Faster: DXC Showcases AI for Quality Control

Quality control is one of the top challenges for manufacturing companies today due to the manual process, which makes it very error-prone. But with AI and computer vision capabilities, manufacturers can implement an automated quality control process that reduces errors, cuts costs, and boosts efficiency. DXC Technology, a global IT service provider, showcased how AI and computer vision help catch defects faster in an Intelligent Boost demonstration at HMI 2024.

Transcript

Russell Duggan-Rees: Hi, I’m Russell Duggan-Rees from DXC. I’m a Chief Technologist within Smart Manufacturing, and I have been part of a team developing artificial intelligence solutions for manufacturing.

Quality control is in the top-three biggest problems for manufacturing companies today. Before artificial intelligence, quality control has been performed by manual workers, very expensive or limited computer vision systems, or not performed at all. Now that artificial intelligence has been introduced to computer vision, it’s enabling manufacturers to affordably and accurately automate their quality control.

At DXC, we’ve invested in end-to-end solutions for quality control, using artificial intelligence in the form of computer vision to ensure our customers can rapidly and affordably use these technologies to automate their quality control, provide quality control at less cost, and provide customers better-quality products.

I’d now like to show you one of our solutions that we have developed for artificial intelligence with computer vision for quality control. Our Chief Architect for that solution, Scott Brodie, will explain a demo that we’ve built for our customers.

(On screen: DXC Intelligent Boost demo)

Scott Brodie: Hi, I am Scott Brodie from DXC, and here we have our Intelligent Boost quality demonstrator. This is just one of our accelerators—we also have operational visibility, smart maintenance, environmental, health and safety, and OEE and KPI reporting. This is just our quality one.

What I’m going to do now is demonstrate this with some wheel hubs. This is an example of good wheel hub, so I’m just going to put this through and show there’s no anomaly on this wheel hub. You’ll see here the green light goes off indicating all is good with this wheel hub.

We’ve got two machine learning models, one on either side, and they detect the anomalies on the wheel hub. In this example we have a bit of oil on the wheel hub. When we put this through now, you’ll see that the red light goes off, indicating its detected anomalies. Both sides detected the anomalies.

I’ll show you another anomaly. Here we have a scratch, there are multiple anomalies on this. We have a scratch and a chip. I’ll put this through now, and we’ll see that once again it’s detected anomalies.

Now on the edge, because there’s edge and the cloud—and on the edge, we are annotating the image that goes up—we can see here, we have the anomalies showing. If we look here, we can see that the oil is on there, it’s given a prediction. And there’s also a drill problem on there. So operators are able to view the anomalies on the product, on the production line.

Many thanks.

About the Author

Christina Cardoza is an Editorial Director for insight.tech. Previously, she was the News Editor of the software development magazine SD Times and IT operations online publication ITOps Times. She received her bachelor’s degree in journalism from Stony Brook University, and has been writing about software development and technology throughout her entire career.

Profile Photo of Christina Cardoza