From Technology Can We Build Tomorrow’s Breakthroughs? I think the answer is a resounding yes though I agree there is reason to be cautious.

As I have discussed before there is at least reasonable evidence that manufacturing in the US is on the upswing and while the charts paint a disappointing picture one has to be a bit careful about what is being measured. If mass production were giving way to boutique build +value add, might one miss it because we are importing what are essentially the bricks and mortar (which we by the way mostly designed or own a big part of) and the customization and final build (with their higher margins and more creative content) is done here?

There is an interesting section in here dealing with the founder of A123 and then with a couple of solar cell manufacturers.  And the author makes some very telling points.  I think they should be emphasized:

1. Unless there is a very fundamental change in a product, improving process technology to produce a that product will not start out producing a cheaper product and unless you can get over the hump of higher cost and lower sales the guy with the bog standard product and highly refined standard process will eat you alive.
2. Controlling one (however important) process or input material does not mean you control the market, a sudden change in market dynamics, possibly one you created, can suddenly pull the prop out from under you and if you only have one prop you are finished.
3. Getting from prototype to production is horribly expensive especially in a mass market (which are almost by definition price sensitive and commoditzed ones) A first article will cost you K, getting that product in front of customers is likely to cost you 2 time K and getting into production 5 times to 10 times as much again, sometimes many times more. 
1. Why you Ask? Because you can do a lot of research for say $1M, that’s enough to support three or four researchers for a year.  But once you have to show it to a customer you have to be able to replicate the work and make either a full scale device and or prove you can do so repetitively or have a process that scales from desktop to garage at least and that usually takes 2 or 3 years or 2 or 3 times as much effort and expense.  The when you go to production you now have to build a factory staff it, train the staff, fill out all the paperwork, pay the lawyers to make sure you’re not doing anything illegal etc, etc.  And you then have to make enough of your product to put on the shelf and most of the time you have to price it at well below cost because the first batches and the smaller batches are much, much, MUCH more expensive than the run of the lot will be later and you cannot charge 10x the expected price.  Selling the first  ???  units at an average of 1/?? their actual cost can eat up a huge amount of money.

The first comment after the article makes the point that established companies have in my way of thinking ‘normalized and processed’ innovation out of their main line business because of the costs.  It is easy to project cost and risk with incremental improvements.   The costs and risks of really new products/processes (disruptive ones) are much more uncertain, and few managers are allowed the latitude to innovate in big, risky ways. 

But this circles back around, does this in fact facilitate the creative destruction that the US industrial base has depended on.  Once large corporations run by bull-headed industrialists did the risky stuff.  When that generation was replaced by the MBA brigade they froze up.  Then the innovations erupted in a series of mid rankers with mavericks at their head or in a series of entrepreneurial start-ups who then took down many of the old guard.  Are we seeing the wake of another change of phase…

Anyway a good article but don’t take it as gloom and doom, its pretty evenly pro as well as con.