A fact that Kim Robertson , the author of over 100 discipline specific training packages, 3 fiction books and articles for CM Trends and various other trade publications from industrial arts to Configuration Management, stated in a recent podcast.
His latest collaboration Configuration Management: Theory, Practice, and Application is
Excerpt from the Podcast:
Joe: One of the parts that jumped out of me was some of the statements right away that you made in the heading of the poor handling of data. I think you mentioned that 30% of the knowledge worker’s times are used for looking for data and even at that, there’s only a 50% success rate.
Kim: Yes, that’s a problem across industry. That one actually came from another book that I got that information from. Basically, we don’t have the data linked. So for example, if you have a subcontract; with the subcontract you have a purchase order, you have a statement of work, you may have some specifications. After that, you have data that they’re going to send it to you which is usually supply chain data lists. Nobody is hooking those together within their product data management or product life cycle management systems. They can tell you every piece part and who the vendor was that goes into the buildup of the final item, but they can’t tell you where the data was, what the receiving’s factual report actually said about the information or much else. That’s very bad when you’re trying to do any type of quality assessment on why things aren’t working the way you thought they would within test before you field something. Or in the case of things like the switch the GM had, where did you go wrong with that piece of it and basically that gets back to one of the premises that if two things don’t look the same, aren’t of the same quality, don’t look alike, don’t keep the same number and we find that that goes on quite a bit. There have been cases of airlines where it’s time to replace an engine and they order a new engine for the jet aircraft, the engine shows up and it doesn’t fit on the wing because they made a change, but they didn’t change the top assembly number.
We have all of those types of activities that we need to take a look at and integrate them together somehow. We have this problem with the Lean-type activities as well. Lean Six Sigma is something that is very popular right now. Lean Six Sigma I believe says you’re going to have two bad parts after every million or so. A couple of years ago, there was a company that ordered a couple million resistors all of the same value from another company and they had a Lean Six Sigma requirement and the company they’d ordered from, the supplier kept saying, we’re going to have to hold off a couple weeks, giving you this last supply data management report. I said well okay, and so then eventually the report came in and there was a shipment of resistors and there were two resistors typed to this note on top of it saying, we didn’t know why you wanted two bad resistors, but it took us eight weeks to find them, since they were working at an Eight Sigma level. So a lot of that, you have to know what your suppliers are capable of before you let your requirements on them because otherwise you may be forcing them to do work and costing you money that you don’t have to spend.
Joe: I think about the data, I think that the inaccuracies that you point out in the data and the lack of cross-references and coordination between all these data, and from a layman’s standpoint it sounds like here we are back to this old file cabinet thing that 80 to 90% of whatever we put in the file cabinet, we never retrieve again. What we did retrieve, a lot of it was inaccurate and though that was in the paper world, is data much better?
Kim: One advantage you had with the paper world was at least you could retrieve it. What we see a lot with the information technology, these activities are that some people make uninformed decisions that the data isn’t required, and so they may dispose of it. We’ve had some cases where I was involved in a program where the decision was made, we’re just paying too much for backing up servers and so we’re not going to back anything after three months. We’ll backup nightly, we’ll back up weekly; we’ll backup monthly. We’ll save three monthlies and then when we save the fourth; we’ll throw the first one away. Lo and behold an entire program’s worth of data went missing and it happened four months or more before it was discovered and the customer was asking questions because the units were still on the field, and nobody could find any of the information. As far as data retention itself goes, I don’t think that we’re much better because the IT organizations and the programmatic aren’t really communicating or understanding if they are talking what the actual requirements are and often its retention for 10 years after disposal of the last unit which with an automobile maybe 40 years from now. On a space asset, some of them would have been up there 35 years. Voyagers have been up there almost 30 years I believe, and it is still going. All of that data is still being retained some place.
Which brings us to a secondary problem is what happens, because the computers are going and involving so quickly. You’re talking about possibly having quantum computers now which would give us something besides buying a recode because you’d end up with 16 possible states for an answer making things not black and white or ones and zeros but shades of gray or Technicolor if you would. So we have some of the missions where we actually archive the computers, the operating systems, the software and everything else because 15 years from now, all of that evolve and there will be no way to actually communicate with a spacecraft that was in orbit. The same thing is going on, on the ground. The last launch of the space transportation system, they were sourcing all 86 boards off of eBay to keep the systems running in order to make that last launch because the hardware was so old, and we have the same thing going on with the data. How do you keep it current and we’ve been struggling with that for a long time. The portable data format or PDF has been a great help with that because PowerPoint and those types of things, when you move them from one server to a lesser cost, storage type of retention sometimes get corrupted and you can’t ever retrieve them again, but PDF’s still seem to be good.
Joe: Configuration management is supposed to help us with all these problems and to me, it sounds good but how does configuration management or what are the keys there to make all these things right as we just talked about?…….
Related Podcast or Transcription: Configuration Management
Lean Sales and Marketing: Learn about using CAP-Do
Lean Engagement Team (More Info)