That being said, the companies I have worked for (admittedly defense not consumer products) were very heavily involved in a "Right first time", "measure twice cut once" style quality system to reduce defective items. Obviously that was safety driven rather than profit driven but it did have a measurable improvement on the amount of rejected tasks or components.

I had an interesting conversation with a chap who ran a machine shop and had a contract to machine exhaust manifold castings for a well known premium brand car. He lost the contract to a Chinese machine shop, the price they charged was hugely attractive but the reject rate was also huge but was still worth it to the end user who had to inspect every component.

Mike

A friend of mine used to run an animal feed mill. Feed is made to specific protein and carbohydrate specs, regardless of actual source of that ingredient, which is how the whole BSE thing happened, one of his ingredients was chopped up Kitkat including foil and wrappers.

If you go round a far eastern market there will be stall full of T-shirts with company logos, I've even seen my own employer's ones. They are the excess made in case of rejects that weren't needed.

That's broadly true, but there are many exceptions. Quality Management has a long history of improvement, but it's not applied equally everywhere.

Modern techniques, like TQM, take a wholistic approach to quality, concentrating on process such that quality is built-in to each stage of manufacture. If anything goes wrong, the process is fixed, proof of the pudding being the ability to turn the fault on and off. The approach aims to eliminate the need for inspectors because they find faults much too late. The manufacturer strives to operate a system that doesn't produce any rejects. Total Quality is typical of high-end products – cars, electronics, white goods, and much else. In the West, it's difficult not to go for low reject rates because labour is so expensive: globalisation means manufacturers can't sell expensive products produced by inefficient staff using low-productivity tooling. What customer are prepared to pay for is an important input to modern quality methods: although customers say they want the best possible, they are rarely prepared to pay for it! Manufacturers are also unwilling to make products that never wear out: they go bust if they do! So manufacturers strike a balance between cost, features and product lifetimes. With low-end goods, it's often cheaper to fix problems by replacing occasional failures than it is to employ Inspectors.

Not everyone operates an aggressive modern quality system, or needs to. Older methods are more appropriate to small production runs and specials. They also work where labour and materials are cheap. Unfortunately, the resulting factory rejects often cause problems when they reach customers. They're sold by both fair means and foul.

Quality systems go wrong! They are particularly vulnerable to human error. Dishonesty, sabotage and laziness as well as mistakes. Interesting example occurred at Sellafield 20 years ago. Sellafield is a British nuclear fuel reprocessing plant, the largest nuclear facility in Europe, and the scene of serious incident in 1957. Very much the sort of installation where I'd expect everyone to be thoroughly on the ball. Not so. After a customer noticed the safety records attached to a batch of fuel were faked, the resulting enquiry revealed incompetent management, a culture of complacency, staff shortages, failure to provide training, inadequate supervision, and widespread falsification of records. All human organisations are prone to this kind of failure and constant vigilance is required.

On the whole I'm impressed at how good most stuff is. I rarely have to reject anything, perhaps because I avoid buying 'too cheap' from unreliable sources and don't have overly high expectations of inexpensive goods, whether shoes or hobby lathes! The duds were all replaced with no fuss.

Dave

Some thirty years ago I was employed by an engineering company, part of a global engineering supplier, to assist in getting the company an ISO Quality Certification. To give you an idea of some of the problems that were commonplace there, the drawing office would produce drawings for parts of the finished product and these would be issued to production, they would give them to machinists to produce parts, the machinists would find errors on the drawings or the parts wouldn’t fit others in the assembly so the machinist would alter the drawing and make the part to fit. The altered drawing would then be retained by the machinist at his workstation to use next time the part was required to be made, no feedback was given to the drawing office so there were now two different drawings for the same part, if and when the drawing office issued a drawing to a sub contractor to produce a batch of these parts when they arrived they wouldn’t fit the product, you can imagine what sort of problems this caused. When I questioned why this was happening I was told well this is the way we have always done it. The company was trying to introduce “just in time” manufacturing and they also wanted to introduce “statistical process control”, this was a recipe for chaos, you can imagine how much scrap was being produced. To give you an idea of how old fashioned the company was there was still the old line shafts running the full length of the machine shop from the days when the machines were belt driven, it only modernised to a degree during the war years when “modern” machines were installed to produce armament equipment. The old values were so firmly entrenched that needless to say the company didn’t get its quality certification and was closed down not long after I left. The product that they were making was a very sound product engineering wise but the quality of its parts meant it gained a reputation for poor workmanship, seeing the situation that we had at that factory it’s not difficult to see why so many incorrect parts end up being sold to customers. Dave W