Issue link: https://iconnect007.uberflip.com/i/1463464
26 SMT007 MAGAZINE I APRIL 2022 bly what has driven us, and it's such an incredible ride right now. Our new facility is in Car- son City, Nevada, and it's significantly bigger than our Santa Clara, California, facility. Carson City will do more production than the one in Santa Clara because building production just isn't economically feasible to do in the Bay Area. It's significantly more expen- sive to build in Santa Clara than Carson City for many reasons, including power, insurance, and so forth. e overhead is a drop in the bucket. We can be much more competitive because we can do production in Carson City and leverage all the tools of Voyager, our soware platform. We are actively extending Voyager to support multiple locations and full-blown production. Matties: Michael, when did you start or bring your Carson City plant online? Kottke: We secured it two years ago, started with the logistics and system integration, and added SMT mid last year. It's just awesome. Matties: You're filling it to capacity, or do you have more room? Kottke: I have so much more room. Matties: Please talk about your soware plat- form and how you decided to be a data-driven factory. Kottke: Data is everything. I tell everybody that we're not a manufacturing company, we're a data collection company that builds printed circuit boards. It's gotten to the point now where there's so much data in Voyager that it drives every decision and predicts solutions. Now we have so much data that we can take it and look at everything from quoting and pricing a job, to equip- ment selection, and defect prediction scenarios where we can flag, "ese com- ponents are something that we see a higher number of defects on. It's in a new design, so how do we pre- vent it?" Matties: Now, your platform considers the work coming in, looks at the components and other factors, then optimizes the flow in your factory for the day, the hour, or the minute. You would be able to shi work around on the fly. Kottke: We do. Voyager changes the schedule hourly. Just this morning, we were trying to fig- ure out how to slip a hot job in for a customer that wasn't on the schedule. ey wanted us to figure out how to build it today, and it's a big job. at's just a normal thing for us, but the rate of change that customers send our way just continues to increase. Matties: With the platform you're using, it's not necessarily a human sitting there trying to fig- ure this out. It's your platform that's doing all the computation, the AI if you will, to optimize your factory. Kottke: at's right. A person double-checks the computations, though, because the plat- form isn't yet smart enough to figure out whether the customer is giving us accurate information. We probably get told every day that something's going to clear, and it doesn't. You wind up putting a job into the schedule, you can't run it, and then you're changing the schedule again.