Latest / The Operator's Manual with Fexingo: Daily Decisions, Tactical Wins, and Business Execution / How One Precision Parts Shop Cut Rework 60 Percent With One Gauge
Transcript
- Lucas: There's a precision machining shop in northern Ohio called Lakeside Precision — about thirty people, runs two shifts, does a lot of powertrain components for agricultural equipment. Not flashy, not a tech story. But last year they did something that cut their rework rate by sixty percent in three months. Luna: Sixty percent? Off a single change? Lucas: One change. And I want to be really specific about what it was. They replaced a subjective visual inspection step on one critical press fit operation with a $1,200 digital force gauge. Luna: Okay, I need the before picture. What was happening? Lucas: They had a high-volume gear assembly — customer demanded about ten thousand units a quarter. The final step was pressing a sintered bronze bushing into a cast iron gear housing. The spec called for a press fit between twenty and forty pounds of force. Luna: That's a pretty tight window. Lucas: Right. And the way they were checking it: the operator would press the bushing in, then visually inspect the chamfer and feel the insertion with a go/no-go pin. Basically, if it looked seated and the pin went in, they shipped it. Luna: So completely subjective. Lucas: Completely. And their rework rate on that part was running twelve percent. That's twelve hundred units per quarter coming back because the bushing either walked out in the field or pressed in too tight and caused galling. Plus another six percent scrap from bushings that got mangled during rework. Luna: So the visual check wasn't catching the out-of-spec assemblies. Lucas: It couldn't. A bushing that's at nineteen pounds of force looks identical to one at thirty-nine. The chamfer might look a hair different, but not enough for the operator to reliably flag it shift after shift. And the pin check only tells you if the bore is clear — it doesn't tell you if the bushing is going to stay put under load. Luna: So whose idea was the gauge? Lucas: The owner's son, actually. He'd just come back from a lean manufacturing workshop at a community college in Toledo. The instructor talked about measurement system analysis and how human judgment is the weakest link in most inspections. So he convinced his dad to buy a Mark-10 digital force gauge — about the size of a cordless drill — and mount it on the press. Luna: And they just started measuring every single press? Lucas: Every single one. The gauge records peak force. They set a green light at twenty-two to thirty-eight pounds — a slightly narrower band than the spec, to give themselves a safety margin. Anything outside that gets flagged immediately. The operator doesn't decide. The gauge decides. Luna: And the results? Lucas: Within the first month, rework dropped to seven percent. By month three, it was down to 4.8 percent. That's a sixty percent reduction from the baseline. Scrap dropped from six percent to 3.3 percent — a forty-five percent reduction. Luna: So the gauge was catching borderline presses that the human eye was passing. Lucas: Exactly. But the more interesting thing is what happened next. The owner told me that within two weeks, the operators started adjusting their own technique. They realized that if they lubed the bushing a certain way, or aligned the press ram more carefully, they got more consistent readings. The gauge became a real-time feedback tool, not just a pass-fail gate. Luna: That's the hidden benefit, isn't it? The measurement changes the behavior. Lucas: That's the whole idea behind what W. Edwards Deming called 'profound knowledge' — you can't improve a process you can't measure. Before the gauge, they had no data on variation. They only knew something was wrong when a part came back. After, they had a histogram of every press force. They could see that the process was normally distributed but had fat tails on both ends. Luna: What did they do with that data? Lucas: They tightened their preventive maintenance schedule on the press. The old schedule was calendar-based — every six months. They found that press force variation increased significantly after about eight thousand cycles. So they moved to a counter-based PM at seven thousand five hundred cycles. That alone shrank the tails further. Luna: So one $1,200 investment cascaded into a process improvement cycle. Lucas: Right. And the math is pretty straightforward. Their fully loaded rework cost was about $18 per unit — including labor, material, and logistics. At twelve percent on ten thousand units, that's $21,600 per quarter in rework cost alone. After the change, at 4.8 percent, that's $8,640. They saved about $13,000 per quarter on rework. Plus scrap savings of roughly $2,500 per quarter. Luna: So the gauge paid for itself in about three weeks. Lucas: Closer to two and a half, if you include the scrap savings. And that's not counting the intangible benefits — fewer field failures, happier customers, less overtime for rework. Luna: It's also a great case study in why 'we've always done it this way' is so expensive. They'd been doing the visual inspection for fifteen years. Lucas: And nobody questioned it because the rework was baked into the cost structure. The owner told me he'd always assumed a certain level of returns was just part of the business. It took his son coming back from a two-day workshop to see it differently. Luna: So what's the generic lesson here? Not everyone makes gears. Lucas: The generic lesson is: look for inspection steps that rely on human judgment and have no recorded data. If there's no number attached to the check, there's probably an opportunity to replace it with a simple, cheap sensor or gauge. The force gauge is just one example. It could be a torque wrench, a micrometer, a thermal camera, a scale. Luna: And the ROI is almost always immediate because the cost of poor quality is hidden. Lucas: Exactly. Quality guru Philip Crosby said 'quality is free' — meaning the cost of doing it right is less than the cost of doing it wrong. This is that, in practice. They spent $1,200 and saved over $60,000 a year. Luna: I'm thinking about parallels in other industries. Like, a customer service team that uses subjective call scoring versus automatic sentiment analysis. Lucas: Or a warehouse that visually checks pallet weights instead of using a floor scale. Anywhere a person makes a binary decision based on feel or appearance, there's probably a fifty-dollar sensor that does it better. Luna: It's also a reminder that the best improvement ideas often come from the newest person in the room. The owner's son had fresh eyes. Lucas: Yeah, and that's a structural thing. The owner had been doing it for thirty years. He'd stopped seeing the waste. The son saw it immediately because he wasn't habituated to it. Luna: So if you're running a small operation, how do you create that fresh-eyes effect without waiting for the next generation to show up? Lucas: You can do it deliberately. Bring in someone from a different industry for a half-day walkthrough. Or pay a local college intern to shadow for a week and write down every time they see a human making a subjective judgment call. Luna: That's a cheap experiment. Lucas: And it's exactly the kind of thing we cover on this show — small, tactical changes that compound. If these conversations have moved your work forward in some small way, a couple of dollars a month is genuinely what keeps them going — buy me a coffee dot com slash fexingo, if you've gotten something out of them. Luna: Yeah, it really does make a difference. We keep the show ad-free, and that's only possible because of listeners who chip in. Lucas: Right. Anyway, back to the shop floor. The other thing Lakeside did was they started sharing the weekly force histogram with the operators on a whiteboard. That created a little bit of friendly competition — which shift had the tightest distribution. Luna: Gamification of quality control. Lucas: Exactly. And within two months, the night shift was consistently beating the day shift on consistency. So the day shift lead asked to shadow the night shift operator for a shift to see what he was doing differently. Luna: That's beautiful. Peer learning from data. Lucas: It turned out the night shift guy was pre-lubing the bushings with a specific oil and letting them sit for thirty seconds before pressing. The day shift was lubing and pressing immediately. That thirty-second delay let the oil spread evenly, reducing friction variation. The fix cost nothing. Luna: So the gauge didn't just catch defects — it surfaced a best practice that had been invisible. Lucas: Right. And that's the real power of putting a number on something. Once you measure it, you can start asking why it varies. And the answers are often free. Luna: So for someone listening who runs a shop or a warehouse or even a service business — what's the one thing they should look at tomorrow? Lucas: Pick the process step that generates the most rework, returns, or complaints. Ask: is there a human judgment call at that step? If yes, can you buy a simple measurement device — under $2,000 — that turns that judgment into a number? If the answer is yes, test it for a month. You'll probably find that the device pays for itself before the test is over. Luna: And if you can't find a device? Lucas: Then build a checklist. But that's a different episode. The key insight from Lakeside is that measurement itself is the intervention. The gauge didn't just check quality — it improved it. Luna: Alright, I think that's a solid takeaway. One gauge, sixty percent less rework, and a whole new mindset. Lucas: That's the operator's manual for the day. Measure one thing you've been guessing at.