Most Equipment Failures Are Process Failures: What $160,000 in Mistakes Taught Me

Most equipment failures are not equipment failures. They are workflow failures. I know that sounds like a bumper sticker. I'm not trying to be clever. I'm the guy at Coherent, the laser company, who has been handling service orders for nine years. I've personally made—and documented—13 significant mistakes, totaling roughly $160,000 in wasted budget. Not because the lasers were bad. Because the process around them was fragile.

People search for “coherent laser company” when they're evaluating suppliers. They ask about power, wavelength, and price. Those are not the questions that will save them money. The question they should ask is: What does the process around the laser look like on the factory floor? If that process is messy, no hardware spec will save them.

The mistake that started my checklist

In my first year, 2017, I made the classic mistake. I swapped a $5,800 laser diode before checking the safety interlock. The system wouldn't fire. I assumed the diode was dead. It wasn't. The interlock was misaligned by half a millimeter. The replacement diode sat on the bench for three weeks, and the customer waited an extra six days.

That error cost $890 in rework plus a one-week delay. I wrote it down. Then I started writing everything down. Over time, that notebook became the team checklist. It wasn't born from theory. It was born from paying the wrong way.

Most buyers focus on the hardware spec and completely miss the process prerequisites: cooling water quality, cleanliness, alignment, operator training, and the checklist that says what to check before pressing start.

Argument 1: Hardware is only half the story

In photonics, coherent beam combining fiber lasers is one of the phrases I still love saying. It's a technique for scaling power while keeping a clean beam. It's elegant. But it only works if the fibers are stable, the thermal environment is controlled, and the combining optics are clean. When beam quality drifts, people blame the laser. Usually, it's a contamination or cooling issue. The hardware didn't fail. The process around it did.

Same thing happens with office equipment. One morning our HP LaserJet kept jamming. I called HP printer support already annoyed, expecting to order a new fuser. The rep asked, “Did you open the rear access door?” I hadn't. There was a tiny scrap of paper stuck in the duplexer. The fuser was fine. If I'd followed the process first, I'd have saved $289 and two days.

The most expensive sentence in any service report is “I assumed.” I've written that sentence myself. Now I ask the team to write down what they know versus what they assume. That one habit caught more issues than any expensive diagnostic tool.

Argument 2: The process is the product

People think reliable equipment creates good results. Actually, good processes create reliable equipment. The causation runs the other way.

Take the Brother laser printer black and white unit in our warehouse. It started leaving gray vertical streaks. The obvious culprit is the drum or toner. The real cause was a dirty corona wire. A $0.10 cleaning fixed it. If we'd replaced the drum, we'd have spent $120 and still had the streak when the new part arrived.

Same lesson in a completely different world. A colleague asked me “how to use a resin 3d printer” last month. He expected a tutorial on slicing settings. The real answer starts with the environment: resin temperature, build plate leveling, and post-processing workflow. If those aren't controlled, no printing profile will save you. The machine is fine. The workflow is wrong.

For reference, the USPS rate for a First-Class Mail letter is $0.73 as of January 2025, according to usps.com/stamps. But the cost of a bad print isn't the stamp. It's the reprint, the delay, and the credibility lost when a quote goes out with the wrong address. Efficiency is where the competitive advantage lives.

Switching my team from “fix it after it breaks” to “check the process before you replace anything” cut our turnaround from 5 days to 2 days. We also added a simple digital log for service orders, and it eliminated the data-entry errors we used to get with paper handovers. That's not bureaucracy. That's capability.

(Should mention: I still use paper checklists for dirty jobs. Digital systems are great, but paper doesn't crash.)

Argument 3: Checklists are memory, not red tape

I know “checklist” sounds boring. But a good checklist is the memory of every mistake your team has already paid for. It's not there to catch people doing bad work. It's there to make sure good work survives on a busy Tuesday. If a checklist is too long, cut it. If it's too short, you'll pay for the missing line.

In Q3 2024, we caught a mis-specified beam delivery test before it went to the customer. The test would have cost $3,200 to redo. It was on the checklist because I had made that exact mistake in September 2022.

We've caught 47 potential errors using this approach in the past 18 months. Maybe 43. I'd have to check the log. Either way, it's more than the zero we caught before. I have never met a company that had too much institutional memory.

The objection: “But sometimes the machine really is broken”

I can already hear the objection: “But sometimes the machine really is broken.” Yes. Sometimes it is. Maybe 20% of the time, based on my logs. I'm not saying never replace anything. I'm saying: before you replace, quantify how much you know about the process. If you can't point to the last time the cooling loop was checked, the filter was changed, or the build plate was leveled, you haven't found a hardware failure. You've found a process gap.

Per FTC guidelines (ftc.gov), claims like “maintenance-free” or “guaranteed output power” need substantiation. I've learned to ask for that substantiation before purchase, not after the first failure. If a vendor won't share the test data, that's a red flag—not because I think they're lying, but because they haven't shown me they can measure their own process.

And yes, some machines are truly dead. I've replaced power supplies, damaged optics, and one very messy resin print bed. I'm not a “process over hardware” absolutist. But I do not believe in paying twice for the same lesson.

The bottom line

I still work at a laser company. I still get excited about good beam quality. But now I know that the most important component in any system is the process around it. The laser, the printer, the resin 3D printer—they all do exactly what their environment allows.

So glad I started documenting my mistakes. I almost didn't. I almost convinced myself that writing things down was for junior engineers. That would have cost a lot more than $160,000.

I should add: none of this is about blaming operators. The fault lies in the process design. If a process allows a critical check to be skipped, sooner or later someone will skip it. That's not a people problem. It's a design problem. And design problems can be fixed.

When I say “process,” I do not mean a corporate flowchart. I mean the steps between “it stopped working” and “we understand why.” A good process makes the right thing the easy thing. In lasers, printers, or 3D printing, the right thing is usually the cheapest thing too. That's why efficiency is competitiveness.