Latest / The Manufacturing Podcast with Fexingo: Factories, Supply Chains, and Industrial Business / How a Japanese Factory Built a 50-Year Maintenance Culture
Transcript
- Lucas: We do a lot of episodes on automation and reshoring and energy efficiency. But today I want to talk about something that sounds boring and is actually the most underrated profit lever in manufacturing. Luna: Maintenance? Because I feel like that's the boring thing that nobody wants to talk about. Lucas: Exactly. Maintenance. But specifically the approach of a Japanese precision-parts factory called Yamazaki Seiko. Three hundred employees, based in Nagoya. They've had zero unplanned downtime for fifty years. Luna: Fifty years? That's almost as long as I've been alive. How is that even possible? Lucas: It's not luck. They practice something called total productive maintenance, or TPM. It's a system where operators are trained to do daily inspections, cleaning, and minor repairs on their own machines. Not a separate maintenance team that swoops in after something breaks. Luna: Right, so the person running the machine is also responsible for keeping it running. That's a big cultural shift from a lot of Western factories where operators just produce and maintenance is someone else's problem. Lucas: Exactly. And Yamazaki Seiko takes it to an extreme. Every operator spends the first fifteen minutes of each shift doing a standardized check: lubricating, listening for abnormal sounds, checking vibration levels. They log everything into a system that's been running on the same database since 1976. Luna: So they have forty years of historical data on every machine? That's incredible. I bet that data lets them predict failures before they happen. Lucas: That's exactly the point. They maintain a vibration archive for each machine — literally thirty years of readings. When a new reading deviates by more than two standard deviations from the historical pattern, they schedule a proactive part replacement during the next planned downtime. They never wait for a breakdown. Luna: So the machine never fails because they replace parts before they fail. That sounds expensive — replacing parts that are still working. Lucas: It would be expensive if you did it blindly. But Yamazaki Seiko's data-driven approach means they replace parts at the optimal point — right before failure probability spikes. Their equipment availability rate is 99.97 percent. That's world class. Compare that to the average US factory, which runs somewhere around 85 to 90 percent availability. Luna: That 10 percent gap in uptime is enormous. If your factory is down even one extra day a month, that's lost revenue you can never recover. Lucas: And it gets worse. A study by the US Department of Energy found that unplanned downtime costs industrial manufacturers roughly fifty billion dollars per year. And that's just in the US. For a single factory, one hour of unplanned downtime can cost anywhere from ten thousand to a hundred thousand dollars depending on the industry. Luna: So spending tens of thousands of dollars on time-based maintenance might actually be cheap compared to the cost of a single breakdown. But can a factory in the US or Europe actually implement this? Isn't there a cultural component? Lucas: There is a cultural component. Japanese manufacturing culture, especially in companies that practice TPM, emphasizes continuous improvement and collective responsibility. Operators take pride in their machines. In many US plants, there's more of a 'produce as fast as possible and move on' mentality. Luna: But that's not insurmountable. I've seen some US auto plants adopt similar programs — Toyota's Georgetown, Kentucky plant, for example, has a strong TPM culture. It can be done, it just requires leadership commitment and training. Lucas: Right. And Yamazaki Seiko didn't build this culture overnight. They started TPM in the early 1970s when the company was much smaller. It took about a decade to fully embed the practices. But once it was in place, the benefits compounded. Their maintenance costs as a percentage of revenue are about half the industry average. Luna: Half the industry average. So they spend less on maintenance overall, even though they replace parts proactively. That's a strong argument for any CFO. Lucas: The key metric to track is overall equipment effectiveness, or OEE. It combines availability, performance, and quality. Yamazaki Seiko's OEE is consistently above 92 percent. For most factories, 85 percent is considered world class. So they're significantly ahead. Luna: Let's talk about the specifics. What does a typical TPM checklist look like for a Yamazaki Seiko operator? I imagine it's not just 'check oil'. Lucas: It's very detailed. For a CNC lathe, the operator checks coolant levels, inspects the chip conveyor for blockages, listens for bearing noise with a stethoscope, takes a vibration reading with a handheld sensor, and logs all data into a tablet. If any reading is outside the green zone, they escalate to a team leader immediately. Luna: So it's not just inspection — it's data collection. And that data is what makes the system predictive rather than just preventive. Luna: And they've been doing this since the 70s. So they have decades of data to calibrate those thresholds. A factory starting today would have to build that history from scratch. Lucas: True, but modern sensors and cloud software can accelerate the process. A new factory can install vibration sensors on every motor, connect them to an analytics platform, and start establishing baselines within weeks. The hardware cost has dropped dramatically. Luna: So the barrier isn't technology anymore. It's the willingness to train operators and change the culture. And that's a management challenge, not an engineering one. Lucas: Right. The CEO of Yamazaki Seiko once said in an interview, quote: 'We don't maintain machines. We maintain the people who maintain machines.' That's the philosophy. If you invest in operator training and give them ownership, the machines will take care of themselves. Luna: That's a great line. I wonder how many factory managers actually believe that versus giving it lip service. Lucas: In my experience, most managers say they believe in preventive maintenance, but when production pressure hits, they skip the checks to hit the numbers. That short-term thinking is exactly why unplanned downtime happens. Yamazaki Seiko treats the morning checks as non-negotiable. No exceptions. Luna: So it's a discipline. And discipline requires leadership from the top. I'm thinking about how this applies to smaller factories with limited capital. Can a 50-person shop adopt TPM? Lucas: Absolutely. In fact, smaller shops can adopt it faster because there are fewer layers of management to convince. One example I know of is a family-owned stamping plant in Ohio with 40 employees. They implemented TPM about five years ago. Their unplanned downtime dropped from 12 percent to under 3 percent in two years. Luna: And what was the key change for them? Was it the training, the data logging, or just the mindset shift? Lucas: The owner told me the biggest change was giving operators the authority to stop the line if they noticed something wrong. Before TPM, operators had to call a supervisor, who might take an hour to decide. Now they stop, fix the issue within minutes, and restart. That alone cut downtime by half. Luna: So empowerment is a lever. Interesting. I think a lot of listeners might be wondering: what's the first step if they want to start a TPM program? Lucas: The first step is to measure your current OEE and unplanned downtime. You need a baseline. Then pick one critical machine — ideally one that's already causing the most problems — and implement TPM on that machine alone. Train the operators, set up the checklists, start collecting data. Prove it works before scaling. Luna: And what about the cost of training? That's often a sticking point for budget-conscious managers. Lucas: Training costs are real but relatively small. The Society of Manufacturing Engineers estimates that a comprehensive TPM training program for a small factory costs about 20,000 to 40,000 dollars for the first year. Compare that to even one major breakdown, which can easily cost 100,000 dollars in lost production and emergency repairs. Luna: So the ROI positive from the first avoided breakdown. That's compelling. But there's also the ongoing cost of the data collection and analysis. Lucas: These days, a basic vibration sensor costs about 200 dollars, and you can connect it to a cloud platform for a few hundred dollars a year. For a machine that's critical to your production, that's trivial. The bigger cost is the time — operators doing checks instead of producing. But Yamazaki Seiko's data shows that the fifteen minutes per shift is more than offset by the elimination of unplanned downtime. Luna: Let's bring it back to the big picture. We're in May 2026, and supply chains are still recovering from disruptions. Any unplanned downtime in a factory sends ripples through the whole chain. TPM isn't just about saving money — it's about resilience. Lucas: Absolutely. If your factory is a link in a global supply chain, every hour of downtime delays deliveries, which can trigger penalties or lost contracts. Yamazaki Seiko's fifty-year record means their customers can count on them. That's a competitive advantage that goes beyond cost. Luna: So the takeaway for our listeners is: TPM is a long-term investment in culture, data, and discipline. It's not a quick fix, but the compounding benefits are enormous. And you can start small. Lucas: Right. And for anyone who wants to dig deeper, the book 'TPM for Every Operator' by the Japan Institute of Plant Maintenance is a practical guide. It lays out the steps in detail. I'd also recommend looking up the case study of Yamazaki Seiko — it's been written up in several lean manufacturing journals. Luna: We'll put some links in the show notes. So Lucas, next episode: are we going back to automation, or do you have another 'boring but brilliant' topic in mind? Lucas: I'm thinking about inventory management — specifically how a Danish hearing aid company reduced their finished goods inventory by 80 percent using a pull system. But that's for next time. Luna: Sounds great. Thanks for listening, everyone. If you found this useful, please share it with a colleague who runs a factory. Lucas: And if you have a topic you'd like us to cover, reach out. We read every message. Until next time.