Latest / The Operator's Manual with Fexingo: Daily Decisions, Tactical Wins, and Business Execution / How One Factory Cut Changeover Time 94 Percent With a Single Metric
Transcript
- Lucas: So there's this metric that most factories track, but almost never in the right way. It's called 'changeover time' — the minutes between the last good piece of one batch and the first good piece of the next. Luna: And I'm guessing the 'right way' involves something other than just counting the minutes and hoping they shrink. Lucas: Exactly. The standard approach is: time it, put a target on it, maybe do a few kaizen events. But the plant we're talking about today — a Japanese auto parts supplier, let's call them Aisin Seiki's smaller cousin — they went from a 90-minute changeover to under six minutes. That's 94 percent reduction. Luna: Wait, 90 minutes down to six? That's not incremental — that's a completely different process. Lucas: It is. And the single metric that made it possible is the split between 'internal setup' and 'external setup' — a distinction from Shigeo Shingo's SMED system. Internal setup is work you can only do when the machine is stopped. External setup is work you can do while the machine is still running. Most factories lump them together. Luna: So if you separate them, suddenly you realize half the 'downtime' work could have been done beforehand. Lucas: Bingo. The plant started by video-recording every changeover. They found that 70 percent of the so-called internal work was actually external — preheating dies, fetching tools, cleaning parts. All of that could be shifted to running time. The first week, they cut changeover from 90 to 40 minutes just by moving tasks. Luna: That's almost too straightforward. Why wasn't anyone doing that already? I mean, the video part alone — just watching yourself work. Lucas: That's the thing — it sounds obvious in hindsight. But in practice, the operators had been doing the same sequence for years. Nobody ever asked 'does this step require the machine to be off?' The metric forced the question. Once they tracked internal versus external separately, the improvement ideas started flowing. Luna: So they had a dashboard showing internal minutes and external minutes. Then what? Lucas: Then they applied the second Shingo principle: convert as much internal setup to external as possible. They built pre-staged tool carts — color-coded, shadow boards so every wrench had a home. They standardized die heights so no shimming was needed. They even installed quick-release clamps that took three seconds instead of threaded bolts that took two minutes. Luna: Those quick-release clamps — that's a capital expense, right? Not every plant can just buy new hardware. Lucas: True. But the plant's total investment was around three thousand dollars for all the carts, boards, and clamps. They saved that in the first month of reduced downtime. The real cost was the discipline to measure internal versus external every single changeover and review the data weekly. Luna: And the floor team — were they on board? I mean, changeovers are often a break time for operators. Suddenly you're telling them to prep tools while the machine is running. Lucas: That's the human side, and it's crucial. The plant manager told me the first reaction was skepticism — 'we've always done it this way, the machine needs to cool down, etc.' But once they showed the video evidence and let the team co-design the new process, it flipped. The operators actually preferred the new system because it was less stressful — no scrambling for tools at the last second. Luna: So the metric itself — internal versus external — became the lingua franca. Everyone could look at the board and see 'we spent 12 minutes on internal today, our target is 8.' That's super concrete. Lucas: And it cascaded. Once internal hit six minutes, they started looking at the external prep: could they stage tools the night before? Could the die be preheated during the previous run? That brought total changeover down to five minutes eventually. The key was that they never stopped asking 'is this step truly internal?' Luna: I wonder how this applies outside manufacturing. Like, think about turning around a conference room between meetings — people shuffle out, new people come in, someone's fumbling with the projector. That's a changeover. Lucas: Right. Or a hospital room turnover between patients. Or even software deployments — the setup time before you can push code. I bet the same metric would expose huge waste. If you time internal versus external for a code deployment, you might find that 40 minutes of 'downtime' is actually just waiting for approvals that could happen in parallel. Luna: And that's the real lesson: the metric isn't just a number, it's a lens. Once you see the split, you can't unsee it. Lucas: Exactly. So if you're running any kind of operation where a handoff occurs — a machine change, a room turnover, a shift change — I'd challenge you to go video a few cycles. Separate internal from external. You'll probably find your own version of that 70 percent waste. Luna: It's funny — we talk a lot on this show about these single-metric transformations. And they always feel like common sense once you hear them. But the hard part is actually stopping and measuring. Lucas: That's the work. And if these conversations have moved your work forward in some small way, there's a simple way to keep them coming — no ads, no sponsors, just the show and a community that supports it. Luna: Yeah, listener support is what keeps The Operator's Manual independent. If you've ever saved a few hours or a few thousand dollars from something you heard here, consider tossing a coffee our way. Lucas: It's buy me a coffee dot com slash fexingo. That's it — no subscription, no recurring thing. Just a way to say 'this helped.' And it genuinely means we can keep doing deep dives on stuff like internal vs. external setup. Luna: So back to that auto parts plant — after they got changeover down to six minutes, what happened to batch sizes? Lucas: That's the multiplier effect. With a six-minute changeover, they could economically run batches of 50 instead of 500. That meant less work-in-progress inventory, faster response to customer orders, and fewer defects — because they caught issues earlier in the run. The plant's overall equipment effectiveness went from 65 percent to 89 percent within six months. Luna: And the metric itself — internal versus external — did that stay relevant? Or did they eventually stop tracking it? Lucas: They still track it every single changeover, six years later. It's on a whiteboard at the machine. The plant manager said the moment you stop measuring a thing, it drifts back. So internal vs. external is now part of the daily standup. They spend exactly 30 seconds on it. Luna: That's the discipline. And it's cheap — a whiteboard and a stopwatch. Lucas: Exactly. So my question for listeners is: where in your operation could you apply an internal/external split? It doesn't have to be a machine. It could be how you process invoices, or how you onboard a new hire, or how you set up a meeting room. The principle is the same. Luna: And if you try it, let us know. We love hearing stories of people applying these ideas. Maybe we'll feature your win in a future episode. Lucas: Alright, that's it for today. Go find your internal setup.