Issue link: https://iconnect007.uberflip.com/i/1108006
APRIL 2019 I FLEX007 MAGAZINE 29 tion. We'd ask, "Why do you want it this way? What if we did this? Okay, that's not going to work. What if we did this instead?" There was back-and-forth communication early on, but Lenthor is the expert on PCB technology—not GraftWorx. We're smart, have a lot of experi- ence, and have done this a couple of times, so we're not completely naïve, but they came up with the proposal for the final platform through a series of talks about possibilities. Matties: That's how collaboration should work. And how important was the price in this model? Flannery: The standard stock and trade of price are always important. But at this stage in the game, the individual piece price is not critical. I don't need to drive this down to the high-vol - ume cost that we expect to get in production, but the platform that you use to first intro- duce your product, get it through regulatory approval, and go into clinical trial with should have a path to low-cost without a major rede- sign. We're a startup. We're cash-poor. Like anything else, every dollar is dear. One of the things I liked about Lenthor's solution was if I want to move to 100,000 units a month or even higher than that, there's a path to get to a low-cost solution without doing a major redesign. In any startup, there's a period where you're trying to do your initial development and alpha and beta units—the stuff that's going to go out for customer evaluation or through reliability and qualification. In that timeframe, it's tempt- ing to take shortcuts and choose something that may be very quick and cheap to build but doesn't scale. The problem with that approach is you get through the regulatory hurdle or your customer's accept ance tests, and then you suddenly find you cannot meet the volume or price point of your customer. You have to go through rede- sign and requalification. Most customers or investors will ride through that with you a second time. The platform we eventually set- tled on had a path to high-volume, low-cost production in its current incarnation. That is very important for us as a medical device company. Matties: That's critically important thinking in the beginning. You're talking as if many com- panies don't start critical thinking early. Is this a common process that people don't follow it? Flannery: More often than you would know, and I'll give you a good example. I was a co- founder of InvenSense. We built the world's first commercial MEMS X/Y gyro and were evaluating foundries. One foundry quoted us a super low number on non-recurring engi - neering (NRE) to build our first device. We thought, "Wow, that's great!" But when we asked, "When we order 1,000 wafers a month from you, what's going to be the per-wafer cost?" That number was very high. Figure 4: Hub and carousel.