Issue link: https://iconnect007.uberflip.com/i/1493797
46 SMT007 MAGAZINE I MARCH 2023 efficient, but what about the data I'm collect- ing? Can I now compare, say, line or plant effi- ciency?" e holistic analysis—the digitiza- tion—of this aspect of the workflow is quite exciting. Right now, we're just deploying the system and quantifying the first-order benefits. But we hear from customers, asking if we can trace back to the root cause and address it in real- time. ere's a lot of potential with this digiti- zation element. Let's look forward to when all these effects start to stabilize. What's your vision on the floor? How will this change staffing for an SMT line? Artificial intelligence is wonderful at the things that humans are traditionally not good at: rote, laborious, repetitive work which, if we were asked to do it for 12 hours, would eventually result in significant inefficiencies. So, why not supplant that work with artificial intelligence and automation, and redirect people to places where they will bring the creativity and the intuition of industry professionals to augment the automation you've introduced? We envision multiple inspection points throughout the workflow. A quality engineer needs to make sense of the inspection trend data and to draw conclusions on how to avoid defects in the first place; the AI's job is to be exceptional at detecting defects. ChatGPT is an interesting analogy; I read an article that said ChatGPT illustrates the dif- ference between writing and thinking. It can produce words that sound nice, but it doesn't understand what you're asking. Now trans- fer that analogy to our industry, where AI will give you all this data, but you need a learned professional to decide what it means. at Darwin AI, VQI Mockup.