Thursday, September 22, 2011

Deliver Us from Beancounters: Learning Curve Edition

Learning Curves and Production Breaks
Dancin' with 'Engineering_Economist' a.k.a. to me as 'EngeCon' again at Defense Tech, and again the DT comment limits prevent material responses (by design I'm sure - which is understandable on a couple of levels) to outrageous posts so here we are again.

This thread is related to my last post, so I don't mind expanding on the point here at all.

(Note: Ignore the Ad Hominem parts of the comments I repeat here, they are only included because I didn't want to be accused of selective editing)

At the DT thread below a post about the Senate Appropriations Committee recommending a two year extension on the current F-35 LRIP production rates, I made the comment:
No doubt in the future, if it is found that the costs of running a fighter factory in first gear for two more years exceeds the total lifecycle cost (or APUC for that matter) impacts that would have been incurred if the conventional alternative (of proceeding with increasing production rates and retrofitting fielded early aircraft) had been followed, that the good Senator (D-Hi) will be the first to admit the strategy was wrong and that he and the other Senators will be the ones to blame for the cost increases. I know that on the other hand, if such a delay does actually save total costs (vs, just playing a convenient ricebowl to plunder for the short term) and has no negative impact on the same costsI will be happy to publically proclaim I was 'wrong'
I won't bother with the preliminary back and forth and the early Ad Hominem but the thread rapidly got to EngeCon posting:
that wasn't even an ad homenim. i asked a question and expressed an opinion. once again the communication process has broken down. also according to learning curve theory it doesn't matter if the lot quantity increases, as long as there is not a production break. so no matter how much you misapply learning theory, you can't blame this on the Senate. the learning effects will not be achieved for a variety of other reasons, such as the unstable production configuration and inevitable obsolescence. there are probably some great introduction to logic and proof classes at your local community college which might give you an appreciation for axioms of math, science, and engineering. i'm sure if we keep repeating the educational process with you, sooner or later something may stick.
The only thing worth noting in this comment was the part about 'Production Break' and 'Learning Curve'. This was funny: A guy trying to 'count coup' while demonstrating a lack of in depth knowledge of the point he was trying to use to refute my argument. I followed with:
Re: Fallacious Ad Hominem – I take the general and to-date unsupported “your obfuscated, hubristic, and biased views” as a personal (not to mention unsupported to the point of nonsensical) attack.
Re: ‘Production Break’ observation. Interesting, but your offhand dismissive use of the term indicates you probably have no idea that a ‘Production Break’(as the term relates to Learning Curves) does not have to be temporal or complete. In fact, your answer reeks of a ‘numerical’ and not ‘operational’ awareness of ‘Learning Curve’ [ I would note that it comes over as arrogance in your own ignorance, but that would mean bringing up “Hybris” again and I’ve noted how that seems to have rankled you some. So I won’t mention it. ] Now go back and try to conceptualize how many of the five categories of ‘Production Breaks’ as identified within the Anderlohr Method are relevant to an operation on the scale of the F-35 and the planned expansion to the manufacturing effort (scope) and speed (production rates). Hint: All of them are relevant.
Re the rest of your poor scratchings: Infantile projection
I don't know why, but I expected EngeCon to at least do a little research on the "Anderlohr Method" before he came back, but nooooooooo:
what is your reference for definition of the term 'production break'? then show how what has happened in the past or in the future indicates that a production break has or will occurr. Congress is not at fault for any production breaks. i understand the frustration in the Appropriations process, but in the F-35's case, Congress has faithfully appropriated billions of dollars into this caper per year, and the program should still be expending dollars that were obligated in past years. the whole experience is ANOTHER lesson in how foolish the concurrency approach is. you need to prove you are ready for production, via a MS C decision, before starting LRIP. what's the status of that DAB to rebaseline the program??
 Aside from the obvious fact that EngeCon hasn't gotten the latest word on Concurrency yet, and aside from attempting to shift the discussion back to the Milestone BS meme he so desperately clings to at the end...there's not much here.  I obliged him with a reference, and would have posted at DT all of what I'm about to post here, except it would have had to be broken into about 6-7 unconnected comments, subject to cherry picking. In 'replay' what I wrote at DT was:
Have you noticed that you often begin with challenging an assertion by asking for ‘support’, then attempt to preemptively refute (poison the well) any support provided in response? And before you have ANY idea what the response will be? Interesting. I assumed you would have just Googled up the ‘Anderlohr Method’. From the second or third hit on my computer:  

RE: MSC/DAB & other formalities. I'll refer readers to our earlier dance :  
What I WANTED to post in addition to the link was the actual categories of "Production Breaks":
My comments in brackets []; feel free to cavil away.
George Anderlohr…. divided all learning lost, by an organization, due to a break in production, into five categories:

"1. Personnel Learning: In this area , the physical loss of personnel, either through regular movement or layoff, must be determined. The company's personnel records can usually furnish evidence on which to establish this learning loss. The percentage of learning lost by the personnel retained on other plant projects must also be ascertained. These people will lose their physical dexterity and familiarity with the product and the momentum of repetition." [From my next door neighbor who works on the F-35 production line, I know that LM is (again) already shuffling key people around into less desirable (read: dirty or at night) work because the production ramp-up is being delayed (again). They are trying to retain the experience and skills of as many less-senior but more ‘F-35 experienced’ line workers as they can--in anticipation of possible layoffs which, given it is a unionized workforce, would otherwise be ‘out the door’ as the more senior ‘protected’ mechanics and electricians are retained.]     
"2. Supervisory Learning: Once again, a percentage of supervisory personnel will be lost as a result of regular movement. Management will make a greater effort to retain this higher caliber personnel, so the physical loss, in the majority of cases, will be far less than in the area of production personnel. However the supervisory personnel retained will lose the overall familiarity with the job so that the guidance they can furnish will be reduced. In addition, because of the loss of production personnel, the supervisor will have no knowledge, so necessary in effective supervision, of the new hires and their individual personalities and capabilities. [We should suspect this will particularly affect those people who were involved in planning the ramp up of production itself, some of whom will start the replan, and others who will now go fight other fires]
"3. Continuity of Productivity: This relates to the physical positioning of the production line, the relationship of one work station to another, and the location of lighting, bins, parts, and tools within the work station. It also includes position adjustment to optimize the individual needs. In addition, a major factor affecting this area is the balance line or the work in process build-up. Of all the elements of learning, the greatest initial loss is suffered in this area." [How much additional production infrastructure had LM and (hundreds of?) suppliers already put into place to support the coming ramp up and will now have to work around it or set it aside?  How much will have to change (do over) if the profile of the ‘ramp up’ has to change (steeper or slower – still a change)? How much more expensive will it be to work through the ramp up in later years, for all the categories and for all the suppliers?]

"4. Methods: This area is least affected by a production break. As long as method sheets [now computerized work instructions- see below] are kept on file, learning can never be completely lost. However, drastic revisions to the method sheets may be required as a result of a change from soft to hard tooling." [ not just ‘soft’ to ‘hard’ tooling – any tooling change driven by scale. BTW I disagree with this being ‘least affected’. The effect is dependent upon the availability of people who understood the intent of the instructions when they were written.]

"5. Special Tooling: New and better tooling is a major contributor to learning. In relating loss in the tooling area, the major factors are wear, physical misplacement and breakage. An additional consideration must be the comparison of the short run or so called soft tooling to long run or hard tooling and the effect of the transition from soft to hard tooling." [Unlike relatively simple manufacturing products, like a microchip or even some F-35 subsystem components, scaling up the tooling for high rate production involves more than just cloning a production line or station. Scaling up the F-35 production will involve reconfiguring and rearranging new tooling. I believe the process of transitioning to full rate tooling has already started with LM subs (remember reading about it in the last year someplace). Will they have to change their processes again to efficiently use the tooling at the lower rate?]

'The definitions presented by Anderlohr have been modified and expanded, since 1969, to accommodate today’s manufacturing environment. For example, some of today’s modern factories operate in a “paperless environment” where method sheets are no longer used. However, these factories normally produce all of their shop instructions on computer files, these computer files sometimes have the same “ability” to get lost as their paper counterparts. Therefore the Methods portion of learning may deal with these computer files (i.e. lost files, changes to files due to new equipment, etc.)." 
Thus it is shown that when the term 'Production Breaks'  is used in reference to 'Learning Curves' within the "Anderlohr Method" (the most widely used application in aerospace as far as I know) it does not just mean cessasion and restart of activity, but applies to any disruption in the current production system that renders prior knowledge ineffective of less effective for future application. As
1. the current F-35 production system had a plan and was executing to that plan, and
2. must now change their execution to meet the next plan's objectives, and
3. that the people and equipment they were putting in place must now change,
new learning will have to begin again or be refreshed.
As the learning curve will now be applied/achieved in later years, the curve can be expected to cost more 'then' than 'now'. If the production rate ramp up under the delayed plan is different from the previous plan, that too will require new knowledge and understanding of the impacts and used to develop the new plan.

Like I said: Cavil away! 

No comments: