Latest / The Operator's Manual with Fexingo: Daily Decisions, Tactical Wins, and Business Execution / How One Toolmaker Cut Setup Time 50 Percent With One Checklist
Transcript
- Lucas: So there's this precision toolmaker in Ohio, about forty machinists, and they cut their CNC setup time by half. Not with new machines, not with software. With one printed checklist. Luna: A checklist. That's it? I mean, I've seen checklists save lives in aviation, but on a shop floor, does it really move the needle? Lucas: That's exactly what the plant manager thought. He'd been doing setups for twenty years, and he could do them in his sleep. But when they actually timed the setups, they found the average was forty-five minutes. And the variation was huge. Some guys took thirty minutes, some took over an hour. Luna: So the checklist was about standardizing the steps? But what's the one thing that actually halved the time? Lucas: The key was in the order of operations. The old way, guys would load the tool, then hunt for the right wrench, then walk back to the cabinet to grab a fixture. Lots of backtracking. The checklist forced a specific sequence: prep all tools and fixtures at the machine before even opening the chuck. That one change cut maybe twelve minutes right there. Luna: So it's like the 'mise en place' of machining. But why did they need a checklist to do that? Seems like common sense. Lucas: Because when you've done something a thousand times, your brain autopilots. You don't see the inefficiencies. And there were two recurring errors: forgetting to zero the probe, and leaving a soft jaw finger-tight. Each of those cost ten minutes to fix. The checklist had a red box for those two items specifically. Luna: The red box. So it's not just a list, it's a visual cue. Did they run into any resistance from the veterans? I'd imagine a guy with twenty years experience doesn't love being told to check a box. Lucas: Oh, absolutely. The plant manager said the first week was brutal. A couple of the older machinists called it 'training wheels.' But they didn't mandate it. They framed it as a pilot, and they asked the veterans to critique it. And the veterans found a few things they didn't like, and they changed them. Luna: So they got buy-in by letting them own it. That's smart. But how did they actually get to fifty percent? That's a big number. Lucas: The fifty percent came from a combination: the order of operations, the red box, and one more thing. They added a 'walk the machine' step. Before starting, you physically walk around the machine, check for loose tools, verify the work light is on. Sounds silly, but it caught three near-misses in the first month. And that walk also let the operator mentally rehearse the setup. Luna: So it's a kind of mental checklist too. But I'm curious—did they measure the time reduction accurately, or was it just a gut feeling? Lucas: They tracked it for sixty days. Average setup time went from forty-five minutes to twenty-two. And the standard deviation dropped from twelve minutes to four. So it wasn't just faster, it was more predictable. Which matters for scheduling. And they found that the checklist caught two errors that would have scrapped a part—one of them was a $2,000 part. Luna: Two thousand bucks. So the checklist paid for itself pretty quickly. But I'm thinking about this from my own work. I've got repetitive processes in my day, and I'm not sure a checklist would make me faster. It might just slow me down. Lucas: That's the classic objection. But here's the thing: the checklist isn't about the steps you know. It's about the steps you forget. The aviation industry figured this out decades ago. The checklist doesn't tell you how to fly; it just makes sure you didn't skip the flaps. Luna: So how do you design a checklist that doesn't feel like busywork? Because I've seen plenty of checklists that are just for compliance, and nobody uses them. Lucas: The toolmaker's approach was: keep it to one page, focus on the critical steps, and make it specific to the machine. Not a generic 'ensure machine is clean.' They wrote things like 'Check that the soft jaws are torqued to 35 foot-pounds.' And they laminated the checklist and hung it right on the machine. No binder, no app. Luna: So it's physical, it's in the moment. That makes sense. But I'm still stuck on the buy-in piece. How did they get the veterans to actually use it every time, not just when the manager was watching? Lucas: They tied it to a small incentive. If the shift had zero setup-related errors for a week, the crew got a free lunch. That was enough to get them to use it consistently. And after a couple of weeks, the veterans admitted that they'd caught themselves almost skipping a step. The checklist became a safety net, not a leash. Luna: You know, this reminds me of something I read about in software development. They have these 'definition of done' checklists. Same idea—make the implicit explicit. But I've seen those fail too. What's the difference? Lucas: I think the difference is the trigger. In software, the checklist is often a form you fill out at the end. Here, the checklist is physically tied to the setup. You can't start the machine without checking the boxes. It's a gating step, not a retrospective. Luna: So it's embedded in the workflow. That's a key takeaway. Alright, I'm sold on the idea. But I want to know, what did they do after the sixty days? Did they just keep using the same checklist, or did they iterate? Lucas: They held a 'checklist review' every month. They looked at any errors that still happened and added a step if needed. But they also removed steps that were redundant. The point is to keep it lean. Their current version is version 14, and it has exactly 18 items. They found that anything over 20 items and people start ignoring it. Luna: That's a nice data point. So what's the one thing our listener should take away from this episode? If they're running a process that's repetitive, what should they do tomorrow? Lucas: I'd say, pick the one process that has the most variation or the most errors. Time it for a week. Then write down the steps as you actually do them, not as they should be done. You'll probably find at least one redundancy and one frequent mistake. Then build a one-page checklist that addresses those two things. Test it for a month, and measure again. Luna: And maybe add a red box for that mistake you always make. I love that. Thanks, Lucas. That's a tactical win we can all use. Lucas: Before we wrap, I want to be honest about something. This show is ad-free, and it's listener-funded. A small group of listeners already chip in a few dollars a month at buy me a coffee dot com slash fexingo, and that's literally what keeps these episodes coming. Luna: Yeah, it's a nice model. No sponsors to please, just us digging into what's actually interesting. If you've gotten something out of these episodes, that's a simple way to keep it going. Lucas: No pressure, though. We're just glad you're listening. Back to the toolmaker—there's one more thing I want to mention. They also started using the same checklist approach for their quality inspections, and they caught a $5,000 error that would have gone out the door. So the principle scales beyond setup. Luna: That's a great example. So it's not just about speed, it's about catching the expensive mistakes. And that's the real ROI. Alright, I think we've got a solid episode. Thanks for listening, and we'll catch you next time.