Caster said:
About 5 years ago, I had a job interview with a firm that wove Kevlar cloth. One of the uses was for body armor. They had a firing range in the building, and every batch got shot with a rifle (among many, many other tests). The city police chief came in to do these tests. My hope is any failure would be really hard to ignore in such a public test.
I think that's an important point. The test was being done on, I presume, newly manufactured vests. What testing is done on a longer scale, to see how the fabric holds up under use?
I noticed how the FDA is reverting its position back to insisting on longer trial studies for drugs, after caving in to pressure from parmaceuticals and others to shorten the wait to bring promising drugs to market. Catastrophic drug effects were found after years of use--"found" means enough failure data was collected on them to reach the red-button point of yanking the product.
Similarly, I wonder how specific was the testing on Bridgestone tires, since the ones that failed were usually on heavy vehicles (Ford Explorer) rear right wheel (when in two wheel drive, the wheel with greatest torque and thus stress/friction on the tire) and a hot environment. (And of course, the lower inflation but not always.)
Modern product development focuses on quick entry into the market. Given a lack of long-term, scientifically controlled data collection, product and process data (also supplier control trouble) is collected via end use feedback. Our latest space shuttle mission was made essentially with fingers crossed. Management, in so many words, said: "We've done all the analyzing we reasonably could and we'll be basically testing our work with this mission." Result? The foam problem was not resolved (new failure data collected in the only way NASA knew how to collect it: under normal operating conditions) but with hyper attention the problem was decided to not ensure failure. The (better controlled than last time) experiment worked: the shuttle came home safely.
So, I wonder how much is understood about why this or that Kevlar woven product fails; if it degrades more aggressively under what environment; if there was a chemical, manufacturing process or other problem that results in failure. How steep the failure curve is--at what time point it is known to become risky enough to pull product.
So while I certainly agree that the company should have tracked performance and recalled the product, not knowing the various factors (including what they knew about how well processes were followed, how deeply or for how long a company is required to test its product for expected performance) I am unsure if that failure alone would necessarily result in ISO decertification.
I guess that is arguably the greatest value in litigation: it's potentially the best way to get such operational data publicized so people can learn from others' mistakes and try to avoid them (I dare to hope).