Issue link: https://iconnect007.uberflip.com/i/1463464
APRIL 2022 I SMT007 MAGAZINE 13 Digitization Matties: What makes it the Factory of the Future? Is it sensors and big data, or is it that the machines are all automated? Kelly: Hands down, the first thing is digitiza- tion; collecting and using data to its fullest, which is easy to say, and harder to do. With respect to automation, though, we need to con- tinue to automate the end-to-end processes, from warehouse all the way to final assembly and test. Matties: But that's something we can do today. Kelly: Yes. Matties: Do we have all we need to do that today? Kelly: Yes. ere is a hierarchy of technolo- gies. For example, people say, "I'll use machine learning or AI." ere is a lot of AI capability right now, especially vision-based AI, but you can't utilize AI without data. Data analytics, consuming data, and moving data are prereq- uisites to do AI and machine learning. It's important to include sustainability. We need to be stronger about sustainability, start- ing with original design and moving all the way through. All too oen this ends up at the bot- tom of priority lists, but it's a critically impor- tant factor that needs priority attention today. When it comes to sustainability, Europe seems to be doing a much better job than North America. What makes a Factory of the Future? Is it automation? Sure. Continued automation is important. But, while many processes are automated, many still are not. Automation improvements in the warehouse are needed, for example. Back-end assembly automa- tion advancements are also needed in test, mechanical assembly, final system box build, and the like. Data is great, but it's also danger- ous. We need secure networks to protect man- ufacturing performance data. Companies are looking for more transparency with quality, yields, and deliveries, but that data needs to be secure between assembler and customer. It would be damaging to pass around quality data and yields, only to have it fall into the wrong hands. Matties: Isn't there a dependency on the design data as well? Kelly: Yes, and that's the biggest one of all. We know everything starts with design. e concept of a digital thread is the nucleus; it explains the data and everything else. Digital thread depends upon having your design cor- rect in the first place; then it gets built up to the system level. You can see how it was made, what's inside, and how it was qualified. It really is the big view. Johnson: Matt, to that point, which is the tail, and which is the dog? Is the design tool envi- ronment driving the manufacturing change- over or the other way around? Kelly: e design data should drive everything, but what continues to happen is that data is peeled off as needed. is has been going on for 50 years; we just peel off the data segments that each entity cares about. It's not a holistic view, even though it comes from the original design. For example, centroid data for compo- nent placement comes from the original board file. A stencil design comes from the top cop- per layer of a design file. The concept of a digital thread is the nucleus; it explains the data and everything else.