Issue link: https://iconnect007.uberflip.com/i/1538540
26 PCB007 MAGAZINE I AUGUST 2025 for this workflow instead of start- ing from scratch. We wanted to own the MES layer and tie it tightly to the business systems. What information matters most when viewing the dashboard? Hendrickson: There are two things: Is our WIP healthy, and is product flowing? I watch the WIP count per area for bottlenecks. I check for when a station isn't producing. For example, with Rout/V-score, they hit goal yes- terday, but their WIP dried up because upstream didn't deliver enough panels. The problem was flow, not productivity. We do the same for the lab queue. For example, we can see the cross-sections they cut yesterday for copper-thick- ness checks vs. final and backup tests. It's all about keeping the line balanced. We now have an organic-con- taminant tester. We use it to ana- lyze the DI water. Because we recycle that water, monitoring organic build-up is critical. At first, we outsourced the testing, but it proved too important, so we brought it in-house for real- time data. J ohn Hendrickson Is the DI-water system perform- ing the way you expected? Hendrickson: We finally hit spec around March last year. First, we had to get the process run- ning; then we learned how to maintain it. GreenSource, our supplier, and Justin Ken- nedy, our engineering man- ager who manages water recycling and the plating line, did an excellent job dialing it in. We now run that system in three shifts, five days a week. We call the group Lab and Water (LAW) because if they see a sudden surge in water use, they have to investigate imme- diately. The balance is fragile; a big draw can throw everything off and shut us down fast. Unlike a traditional once-through system, we can't just bring in fresh water and dump the rest. So, the system isn't a bottleneck? Hendrickson: No, but is it sized for our 2030 production tar- gets? Maybe not. We used aver- age loads when we designed it; spikes weren't on our radar. Now we understand those spikes, and that's why we staff a third shift. The recycling plant probably needs to grow to give us more headroom and flexibility. What have you seen in terms of measurable outcomes— things like cost, speed, and control—by bringing PCB production in-house? Hendrickson: There has def- initely been quality improve- ment across our factories. As for cost, it's tracking right where we expected it to be, which is good. But much of the value lies in the intangibles. I spent 20 years designing these boards, and let me tell you, designing them was a lot easier than build- ing them day in and day out. The little things add up. For example, we looked at one of my designs from back in 2003, where a seemingly minor rout- ing decision ended up impacting throughput. Just that one small routing path made a difference. Now, we ask how we can bet- ter educate our design teams. Should we be more challenging in our design reviews? Should we expand certain features? There are so many opportu- nities to improve to ultimately make the product lower cost and more reliable. There are many subtle, intangible factors that we need to start addressing by working more closely with our partners in R&D and manufactur- ing. Doing so can help us save cycle time, reduce costs, and ulti- mately improve overall reliability. Jessi Hall: One of the other mea- surable outcomes we've seen is when we've run into challenges with our printed circuit board J essi Hall