The Inspection Passed. The Product Still Failed.

Too often, AI ‘human in the loop’ describes a position in the workflow, not control over it

The vision system scanned the part and logged a pass. The SPC chart stayed inside its control limits. The line kept moving.

Three weeks later, the customer complaint arrived. The defect had been there the whole time, sitting just inside the algorithm’s confidence threshold, close enough to spec that the model called it good.

Nobody did anything wrong. The camera worked. The model performed within its trained tolerance. The dashboard was green. The product still failed.

This is the accountability problem now facing quality organizations that have moved AI into inspection, supplier scoring, and CAPA workflows. The technology isn’t the issue. Most of these systems perform exactly as designed. The issue is that almost nobody redesigned who owns the decision once the machine starts making it.

That isn’t a technology problem. It’s an operating model problem.

Presence is not authority

Most quality organizations already have a human somewhere in the AI-assisted process. A quality engineer reviews the exception report. A line supervisor gets a notification when the model flags a borderline call. On paper, that looks like a human in the loop.

…

Want to continue?

By logging in you agree to receive communication from Quality Digest. Privacy Policy.

Create a FREE account

Forgot My Password

Comments

AI inspection

This article brings up an extremely important point consideration -- the relationship of responsibility between the machine/software and the human. Many of these systems are justified on the false understanding that they will be able to take over human decision making or review tasks and thus eliminate labor cost. The article makes it clear that in class 2 and 3 decisions, humans must remain involved and engaged so that the organization is protected from the imprecision that still remains in vision systems, programming and AI statistical approximations.

I spent more time reading than you spent writing. You're welcome

"That isn’t a technology problem. It’s an operating model problem."

This isn't an article on Quality Digest. It's AI slop.

The author's LLM-of-choice contrives a scenario in which "nobody did anything wrong," but somehow the manufacturing process was operating right up against one of its specification limits, even within the uncertainty of the measurement system, and also the process was operating within its natural process limits. So you have a process that is in statistical control, but the specification limits are within the natural process limits... because "nobody did anything wrong," right?

This is Steve Carrell in 40-Year-Old Virgin talking about how "it feels like... a bag of sand." Or the "three fingers" scene from Inglourious Basterds.

First thing that someone did wrong was fail to read Don Wheeler's articles from June and July of 2010: "Is the Part in Spec?" and "Where Do Manufacturing Specifications Come From?" That's where you'll find the needed theory, and it applies whether you use AI tools or not. Ignorance of it will hurt you, whether you use AI tools or not.