"I don't see it," I say.
"It's right there," says the agitated part inspector, pointing with increasing impatience as I strain to find the defect they're trying to show me.
"Wow — is that something you can't ship?" I ask, squinting at a nearly invisible imperfection.
"Yes. Now let me log it in."
I've spent time in plants where JD Power Initial Quality was the goal. That award looks at issues reported by owners in the first 90 days of ownership — many cosmetic, some would argue minor, but absolutely real to the customer paying for a premium product. I was at a GM plant that won it multiple times. It was a genuine source of pride. And behind that pride was rigorous, unrelenting visual inspection.
How visual defect tracking works today — and where it breaks down
In most plants, visual defect tracking looks like this: a grid tally sheet with defect locations on one axis and defect types (scratch, dent, rip, bleed-through, contamination) on the other. Inspectors tally as they go. Some use coloured pins on a diagram to show where defects cluster. It's manual, it's physical, and it's been done the same way for decades.
Parts that fail — defects that can't be repaired at the primary station — get a tag describing the problem and go to a repair rack. The secondary inspector then has to decipher the tag, find the defect on the part, and repair it. Finding the defect can take longer than the repair itself. And repair racks move around. Getting a true picture of how much product is sitting in repair at any given moment is sometimes a full day of walking the floor.
The data from this process — all those tally marks — is supposed to drive quality improvement. In practice, it drives end-of-month report writing. The insight disappears with the paper.
What it looks like with a digital system
Now imagine instead: the primary inspector scans the serial number of the part. Model-specific images of that part appear on a touchscreen — front, back, key zones labeled. The inspector taps the location where the defect is found and selects the defect type from a concise list. The data is captured instantly, tied to that exact unit.
Defects that can be repaired at the primary station are marked repaired on the spot. Defects that need secondary attention remain open. When the part reaches the secondary inspector, they scan the same serial number — and the screen shows them exactly where the defect is and what it is. No deciphering. No hunting. The time savings are real.
At any moment, a supervisor can see how much product is in the repair queue, what defect types are driving it, and which lines or operators are generating the most rework. Pareto charts update in real time.
The poka-yoke that makes it complete
Extend the system one step further: when parts are scanned into the shipping container, the system checks the inspection history. A part with an open defect — one that hasn't been repaired or formally dispositioned — cannot receive a shipping label. And you cannot ship without a label.
The system ties inspection outcomes directly to shipping. Not as a manual check someone might skip — as a hard gate that cannot be bypassed.
This is a poka-yoke in the truest sense. Not a reminder. Not a warning. An actual prevention. Defective parts don't ship because the workflow won't allow it — not because an inspector caught it at the last second.
This is what 10in6 Visual Inspection delivers: structured defect capture, real-time repair queue visibility, Pareto data by line and operator, and a shipping gate that closes the loop. It can be live in your plant faster than you'd expect.