One of our regular readers, Brad Pritts, sent me a note about yesterday’s article. I thought it was a great follow up. Here it is.

—

I’ve been working with ISO TS 16949 since its inception, and with its predecessor, QS-9000. I’d have a few comments for those unfamiliar with it.

1) Why did it come about?

As mentioned, the auto industry is highly dependent on supplier quality; they’ve had their own approaches for SQA since the late 1970’s. Besides the difficulty of change, think about the complexity due to the numbers of parts involved in a car. Typical assembly plant gets deliveries of 2,000 – 4,000 part numbers to build the cars. This number counts complex assemblies such as engine, transmission, and axle as 1 part each, so the total number of components is 10,000 to 20,000. If any one is defective the whole car is defective to some extent (an oversimplification, but not by much.) If each supplier has a failure rate of 1 part per thousand, then the typical car will have 10 – 20 defects. So we aim for perfection, with actual performance (ideally) in the parts per million defective range.

2) How is it better than ISO 9001?

A. Many auto industry specific additions, as Renaud listed in the article; mostly good practices. The “Core Tools” in particular provide very specific and prescriptive methods for quality planning.

B. A strict administrative/accreditation process that weeds out poor quality auditors and shady registrars. (Unfortunately it also discourages some good auditors, but that’s another story!)

But there are also some disadvantages/ problems, such as:

C. A vaguely defined, clumsy (in my opinion) method of implementing the “process approach”. I have no dispute that the ISO 9001 standard requires the adoption of a process approach but the TS oversight people have cobbled up what I consider a poor approach for process management. Furthermore, they have never codified it, so the only
way to learn it is from the auditors.

D. A few requirements which are impractical for some suppliers but are nonetheless mandatory, with little room for auditor judgement to consider them as not applicable. One, for example, is the requirement for “predictive” maintenance techniques. These are great for some industries (e.g. vibration studies for large complex machinery) but not really helpful for other simpler processes such as stamping.

E. An insistence on prioritizing audit time to the specific products made for the TS sponsoring companies (GM, Ford, Chrysler, Renault, VW, etc,. but NOT Honda, Toyota, etc.) This results in odd audit schedules and practices for companies who have limited business with the sponsoring companies; the same GM or VW job may get audited again and again while the Honda job never does. Quite contrary to the normal spirit of random auditing.

I tell quality professionals from other industries that they should study the auto industry and TS 16949 as a role model. Many of its provisions are great; and even the ones that are lousy can always serve as bad examples!