Issue link: https://iconnect007.uberflip.com/i/1479191
18 PCB007 MAGAZINE I SEPTEMBER 2022 ning perspective, in the U.S., you typically find that one machine will do many different pro- cesses. ere are different recipes associated with different processes. If you have electroless copper, desmear, or something like that, those aren't just for desmearing the hole or putting electroless copper on a through-hole. ey do blind vias, maybe a button plate, or a cap plate over epoxy via fill. You also have pattern vs. panel applica- tions, seed layer management, and it gets quite sophisticated. You can build products in the U.S. market that take 400 steps. is is typi- cally not the case in production shops where they streamline things, have bigger lines, put more steps into the line, and standardize. In the U.S., you could best accomplish this all through a good product lifecycle management (PLM) system. ey're very cost effective, and you can use the PLM to architect your bills of process: how the recipes work, how things interact. You then have a nice framework for how to automate your factory. It's an excellent soware tool with a very low cost to use as a framework. I would also complement this with a DSM (design-structure-matrix) optimization of the bill of process (BOP). Without this disciplined approach, you auto- mate based upon all the evolved operations, but the capital efficiency and ROI are less if you do it in this way over the long term. You can't just say, "I'm going to have these stan- dard work cells, and put the same loader and unloader concept on every single one, because it's too agile for this." Instead, you must say, "Every machine is different, and I may need to have a different solution for each one, a differ- ent system architecture, and how I'm going to manage this system to maximize ROI." Don't do standardization for the sake of standardiza- tion even if it looks good on paper on the sys- tem level, but focus more on the subsystem level to get the savings. Each work cell has a dis- tinct system. You take them one by one, based upon the return on investment. First, tie it into projects with clients and that's the approach. Stepinski: Look at it this way: Why are Europe and Asia automated and the United States is not? Why is that? Matties: What are your thoughts about that? Stepinski: You have more production in those regions, so you get economies to scale from the automation, and the competitive landscape is based upon who can make the widget cheaper. e U.S. outsourced all its production. Europe outsourced too, but there are some German shops still doing significant production for the automotive market. In those shops, you see the automation. Other shops in Europe, even if they're not doing that mass production, see the automation in Germany, and they adopt those practices because they have something to ref- erence. It's easy to visit the German shops, see how it's done, and learn from it. In the U.S., we don't have these reference accounts. Everybody locks their doors and they're not as friendly. I find the EIPC conference, for example, to be a much friendlier place than perhaps IPC APEX EXPO when fab suppliers are talking to each other. I see a lot more cooperation on the European side than on the U.S. side, for instance. e U.S. and Europe are analo- gous in a lot of ways. Europeans outsourced, but they retained their automotive, and you saw advancements. In the U.S., we outsourced everything, and many of the shops that had automation closed up. In Asia, they went full-bore automation, because this is where all the production is. You get scale economies from automation; you compete better when you add automation. If you're in an agile situation, you don't have the same automation strategy as a volume production shop. It has to be semi-automatic because you won't set up your process to build one part number all day. In this instance, you treat everything like a work cell, and the focus is this: What products go through what work cell, and when/how. From an operational plan-