Last February, I answered a phone call that started the way a lot of quality complaints do: Your Mando struts don't fit. We're stopping the line.
I'm the quality compliance manager at Mando, a company that makes stamped metal parts, dies, forged parts, and CNC-machined components for automotive customers. In my role, I review every lot before it ships—roughly 200 unique part numbers a year. I've been doing this for over four years, and I've signed off on parts that kept me awake, rejected first articles that annoyed suppliers, and stood on a plant floor with a caliper at 11 p.m. more than once.
That day, the issue wasn't a complicated CMM measurement. It was simpler and uglier.
Why the line stopped
The customer had a large order of Mando struts on the bench. They were a new revision of a suspension part we'd been making for two years. The mounting bolt pattern on one end was drawn as a 4-hole pattern with a 97 mm bolt circle. The actual part measured 95.8 mm on two of the holes. That's 1.2 mm off. On a suspension component, that's not a cosmetic issue—it changes the clamping load and can create a clunk as soon as the vehicle rolls out of the dealership.
I know that clunk because I've heard it in a test vehicle. It's the kind of noise that gets a warranty claim before anyone knows the strut is the cause.
So we rejected the batch. All 600 pieces.
What the Mando login records showed
The first thing I did was pull up the Mando login portal. Our internal system links each serial number to the heat lot, the tooling ID, and the inspection records. It's not magic—it's a database behind a login. But it works. I traced the batch to a progressive die line we run for suspension products.
The production log showed an entry I almost missed: a maintenance action at 10:37 p.m. on the night before. A die insert had been replaced. The tooling tech signed off, and the line restarted at 11:20 p.m. But there was no first-article check after maintenance.
In my first year, I made the classic specification error: I approved a preventive maintenance schedule without requiring a post-maintenance first-article check. I learned that lesson the hard way when this batch came back. It cost us a $22,000 rework, two weeks of delivery delay, and the customer's patience. I'm not 100% sure the insert swap caused the misalignment, but we couldn't prove it wasn't the cause. That's the real problem.
What a sensor diagram taught me
Stories about bolt patterns on struts might feel far from engine diagnostics. But they're not. Most of the searches I see for where is mass air flow sensor are really about the same thing: people understand what the component does, but they don't know where it goes or how it's oriented.
A crankshaft position sensor diagram looks simple—a small sensor mounted near the crank, usually behind the harmonic balancer or near the flywheel. But the diagram only helps if you know the air gap and the wiring route. If it's mounted upside down, or the harness sits too close to a heat shield, you get intermittent no-starts. The part is right. The location is wrong.
After the strut incident, I started requiring our engineers to include location photos in every first-article report. It sounds basic. But if a technician has to search where is mass air flow sensor for a common V6, imagine what they do with a custom stamped bracket for a prototype.
The Honeywell smart thermostat app lesson
There's a parallel in my own house. Earlier this month, I set a new HVAC schedule, and the Honeywell smart thermostat app kept showing the room temperature six degrees lower than it actually felt. I checked the WiFi, the wiring, and the app settings. The problem was simpler: the room sensor had been knocked off its shelf by a window blind. The app wasn't lying. It was reporting what a misplaced sensor saw.
That's exactly how a car acts when a mass air flow sensor is mounted too close to a kinked intake boot. The sensor sees the number it should, just not the number the engine needs. The fundamentals haven't changed: sensors don't lie; locations do.
What changed after the rework
We fixed the strut batch, replaced the insert, and ran a 100% dimensional check until the process got back to an acceptable Cpk. The part has been stable since. But the bigger change was procedural: no maintenance event on a stamping tool closes without a first-article inspection on the first piece after restart. That rule cost us a $22,000 lesson to learn.
I also added a requirement that every new part and new supplier must include location photos for critical characteristics. Some engineers still think it's overkill. I usually respond by showing them the 1.2 mm gap photo and asking if they want to bet a production line on it.
What was best practice in 2020 doesn't necessarily apply in 2025. Five years ago, some customers accepted static PPAP files without location photos. Now they want digital records tied to serial numbers. The execution has changed, but the fundamentals haven't: know what the part is supposed to do, prove what you made, and make sure it goes in the right place.
This all happened in early 2025. Mando's internal systems and part revisions change, so verify current specs before ordering tooling or replacements. I'm not giving engineering advice here—just sharing a process lesson.
And one more thing: I'm careful about the word guarantee. If we claim a Mando part is precision-stamped and meets OEM spec, we need evidence. FTC advertising guidelines require substantiation for claims, and the spirit of that rule is one I agree with. If you can't show the measurement, don't make the claim.