Ok, for the sake of discussion, what is an acceptable failure rate for anything?
For a brief discussion of failure rates, look here. Complex isn't it?
.5 = 50%...
the divorce rate in the US
.25 - .1 =
is the rate of miscarriages
.005 =
chance your house will be struck by lightening, 1 in 200
.0000036 =
chance you will be struck by lightening, 1 in 280000 in USA.
.0000034 = 3.4 failures in 1000000 (a million) which is the
much touted six sigma rate for a
single defective manufacturing part per million opportunities.
GM has linked 54 accidents and 13 deaths to faulty ignitions,
according to USA Today.
.0000027= 54 accidents in 20 million vehicles (20,000,000) = .0000027
This means that GM has achieved better than six sigma in its manufacturing process. I'll leave it to the statistical wizards and six sigma guru's of TBN to determine if a key switch is a manufacturing opportunity, or is the entire vehicle a manufacturing opportunity?
SO, AGAIN, WHAT IS THE ACCEPTABLE FAILURE RATE, IN A REAL WORLD WHERE STUFF HAPPENS??
To have a valid response, you need to include a number.
My vote is that six sigma, .0000034 failures per DPMO, is adequate, in a real and practical world, requiring no change in part or recall.
So, what is your number and how did you achieve it??? Please support how your number is achievable, in a real world, with references, as I have done.
No number and no supporting references = no reason to listen, it's just puffery.
Just for the sake of discussion, and recognizing that people suffering death or accident, of any kind, from any source, have been grievously harmed and that all possible efforts should be made to avoid such future harm.
Any six sigma green/black/master black belts/champions or executive leaders out there who wish to identify themselves and comment?
This is
GM'S latest comment on the subject.