Jeep and Quality in Product Development

A Curious Post on Jalopnik

Last week, I saw an article on Jalopnik ?about the launch of Jeep’s new flagship SUV, the Grand Wagoneer. It seems that a pre-production vehicle had a serious cosmetic flaw, which you can see clearly in the Jalopnik article.

Below the rear window glass on the liftgate, the word “WAGONEER” is spelled out in nice chrome letters, each letter individually mounted on the liftgate above the license plate depression. The problem is simple: the letters aren’t even close to being parallel with the bottom of the glass or the trim just below this brand identity feature. Not only that, but the letters also aren’t just in a non-parallel line, they are mounted in some kind of an uneven curve.

It might be a nice math assignment for someone to derive the equation for the curve this feature makes.

As the article notes, however, Jeep (or Stellantis) acknowledged that this had occurred, but had been found and corrected, and no vehicles with this flaw were shipped to customers. In addition, Jeep then explained that locating fixtures and a vision system had subsequently been installed to insure that this trim was properly installed in the future.

Jeep then added, “The plant has a robust vehicle audit process that ensures every vehicle leaving the plant reflects the premium quality of the brand.”

No doubt!

What’s Going On At Stellantis That Led To This?

This might sound like just another routine early-build glitch that arose, but it seems to say something much deeper and significant about product development processes at Stellantis.

In the past, I’ve posted two articles about this in the Lighthouse library I’m building on LinkedIn. These two articles bear directly on what we can see in this story about the launch of the Grand Wagoneer.

The first article is about the difference between prevention and detection controls in any quality system, and why prevention controls are extremely important. If you aren’t really sure why prevention controls are so important, you should read this article.

But if you don’t want to take the time to read the article, consider this: the most inspected item in human history, a single component with no moving parts, was repeatedly checked for defects using an almost unbelievable array of extraordinarily sophisticated and complex tools. The final inspection was signed-off by dozens of highly qualified professionals.

Despite that level of inspection, the product was released for use. And it was still defective, and defective in a most fundamental way!

That component was the mirror for the Hubble Telescope.

Everyone knows that “100% inspection isn’t 100% effective.” So why do so many manufacturing control plans rely overwhelmingly on detection controls—inspection looking for defects? Or as Jeep said, “a robust vehicle audit process.”

This demonstrates something I’ve told people over and over again—even though 100% inspection is absolutely not 100% effective, almost every factory manager I’ve ever met behaves as if it IS 100% effective.

The question that should have been asked is why weren’t the fixtures and vision system part of the pre-launch control plan? Was it an oversight? Was it considered and not used to save money? Or was the whole thing rushed into production without sufficient thinking and planning?

I don’t know the answer to this question, but if I were an executive at Stellantis, I would certainly want to know the answer, and that’s because the answer is the real root cause of this problem! If that root cause isn’t discovered and corrected, this will happen again. Moreover, the idea that a “robust vehicle audit process” is the key to vehicle quality will be perpetuated, and without a single solitary doubt, will result in vehicles being shipped with defects.

Maybe that’s why Jeep’s Grand Cherokee rated 6th out of 7th in J.D. Power’s “Best Midsize SUVs of 2021” ratings.

However, I’m afraid this will be blamed on “poor workmanship” and that will be the end of it.

Could It Be Poor PFMEA Methods?

This strikes me as a likely root cause, but even if reasonably sound methods are used for Process Failure Modes & Effects Analysis, the results aren’t always applied meticulously. I dealt with this just a few weeks ago in another Lighthouse post “Must Control Plans Contain Every Control in a Process Failure Modes & Effects (PFMEA)?”

It’s possible the fixture-and-vision system was considered and not used. It may even have been entered on the PFMEA and didn’t make it into the Control Plan for the final trim assembly process. After all, fixture-and-vision isn’t new or innovative; it’s been used for several years in many assembly plants.

However, in my experience—and I don’t know what went on at Stellantis on this vehicle development program—it’s much more likely that the PFMEA was carried out in a perfunctory manner, with far more concern about being “compliant” with the AIAG-VDA FMEA Guideline than concern for a best-in-class quality SUV. It’s also probable that part of that cursory effort was an over-reliance on software tools to crank out the PFMEA and subsequent Control Plan.

All Of This Makes Me A Bit Crazy

As I write this note, I’m 71 years old, and I’ve been watching this kind of half-baked quality management in factories all over the world for my entire adult life. It never made sense to me, not even when I was a young engineer. All the while, I was told over and over and over that “it’s good enough” or that “you’re just not being practical” or given some other excuse. Sometimes, I'd just get blank looks when I raised a question about a poor choice.

I’m starting to believe that I won’t live long enough to see good practices for quality management adopted on a wide-scale basis in most manufacturing facilities. And that’s frustrating because I know that addressing this is well within the grasp of almost any organization.

We can do better.