Issue link: https://iconnect007.uberflip.com/i/1428512
68 PCB007 MAGAZINE I NOVEMBER 2021 bility and your tight process control. at's got to be a big measure of ROI. It's not just your yield, but it's also the market opportunity that you create by putting yourself in an advanced processing position. You saw that, no doubt. Stepinski: Yes. ese types of techniques are very common in other industries, like tuning things on the fly based on sensor measurements and improving your modeling. It's just some- thing that really hasn't been in our market, for whatever reason. It seems like it's a low-mar- gin business; there's not enough R&D fund- ing. is is where we ended up as an industry. I mean, even the pizza delivery industry is more advanced. ey have digital tools to tell you when the pizza will be done. We're just start- ing this stuff in PCB. Matties: If we were to look at a brownfield site, where should the fabricator start? Is there a self-audit or some next step that they need to take to understand where they're at, what the opportunities are? Stepinski: It comes down to looking at what you are and are not controlling, what data you have to use, and how you can connect what you have. at's the first step. I think you can rely a lot on the suppliers for this. If you just convened with your suppliers and said, "We want to have a more digital factory. We want to move forward into the 2020s here and join the digital revolution. A little late, but we're here. How can you guys help us?" at's the first step, because that's where all the research has been done right now, at the supplier's side. You'll find the AOI equipment actually uses artificial intelligence. ey have unsupervised learning algorithms that group things together. You can say, "ere are 500 pictures that have one speck of black stuff on them that appears to be oxidation. It's not a real defect." It groups them together, so you can say, "I'm going to not inspect this family. I'm not going to ver- ify this family of defects, but I am going to ver- ify the 11 that look like they're nicks and see if they are really nicks." ese tools are already in place to some extent, but do the shops optimize this and develop their procedures? I think the suppliers are further ahead than the board shops. Johnson: What do you see as the tipping point for the industry? Stepinski: I don't think that there have ever been big tipping points in our industry. I think it's always been quite slow. But what you see now are some of the bigger companies based out of Asia who are going in this direction. ere's a lot of Industry 4.0 going on there, but it's mass production oriented, for the most part. In Europe, you see some R&D develop- ing these processes for higher mix, but a lit- tle slower. We see some market growth in Europe, which may result in more investment as well, to improve this. Once you have these case studies, then the suppliers are the ones who distribute the knowledge. ey don't take ideas from their clients. ey just say, "Hey, we did a big project." Everybody hears, "Hey, this company did this." You have something that we could use, and there's a little more background to it, and they can help you develop it. ey give you a little head start. I think that is what will happen, but it will be a slow osmosis into the market of these things. My recommendation is not to wait for the slow osmosis. Get ahead of it. Take some classes on Coursera, go to the Hanover Fair in April (or I mean, even the pizza delivery industry is more advanced. They have digital tools to tell you when the pizza will be done.