Tuesday, July 1, 2008

recent reading: zero quality assurance, getting things done

i have been trying to read Shigeo Shingo's epochal book 'Zero Quality Assurance' for the last few weeks. not much progress - and Shingo is not to blame for that, i am getting crushed by the urgent and not giving enough time for the important!

the thing that hits you right in the face is the amount of clarity this guy's thinking has.. i hv always believed (and so says Eliyahu Goldratt in one of his books) that the best argument is one where the listener responds saying this is so common-sense, and also has an @aha, so that is the explanation" feel to it.. the same thing i felt when i was reading Shingo.

key thoughts from whatever i have covered of the book till now:

1. in quality assurance, assurance is the deity and statistical methods is the temple.. but we forget the deity and are over-awed by the temple. statistics is glamourous.. but is not needed. if you can do 100% in-process inspection, SQC is not needed.. and with 100% inspection and poka-yoke (mistake proofing) zero defects is possible - and has been achieved (he gives examples of a mitsubhishi or matsushita plant where this was achieved)

2. work has 2 perspective - that of operations (functional organisation) and that of process (or product flow). often the operations view dominates, because that is very visible day to day. but we should give more focus to the product view - if improvements are to be effective (this is the global vs local optima and throughput focus that Goldratt too tries to bring in his Goal and theory of constraints).

3. all operations can be classified as either Work itself, wait, transportation or inspection. and work itself can be classified as 'set up for work' and work itself. ( i thought he has not added 'rework' as a category of work too). and he says that all else other than work itself (excluding set up) is a non-value add... this, i have been using in our value stream mapping exercises extensively.. we never had an objective framework for judging whether an element value adds or not and i found this classification extremely useful..

more soon!

cheers, suresh

No comments: