After taking a brief hiatus due to work concerns, Dan Schrimpsher of Space Pragmatism just posted an interesting reply to my last post.
Dan briefly discusses the idea of change in development processes, and brings up the concept of what is called frequently called “agile” development. In his own words:
Agile development was the created with the idea that change isn’t bad. It isn’t good, it just is. It will happen, so why not design your system from the beginning to be fast, light, and flexible.
I think this is a very important concept myself. Many of the grand plans I see coming out of NASA, space enthusiasts, and even many space companies seem to imply a static world. One where nobody else does anything in space other than buy your goods or services, or fly on your vehicles. This is related to my whole rant against monocultures.
Reality is quite different from these static worlds. Not only are there other companies, groups, countries, etc pursuing various space projects of their own, there are also many non-space related groups pursuing their own techologies, markets, etc. that all have to interact together. Michael Mealling brought up the concept of Value Networks during the RTM conference. The interesting thing is that the more you think about it, these networks are constantly evolving and adapting. The history of the civilization is one of constant change. Of technologies, markets, and free association interacting in sometimes quite rapid and unexpected ways. As Dan puts it, change isn’t good or bad, it just is. Whether you’re NASA or a private company, you’ll get just as screwed if you ignore change.
While I’m sure there are all sorts of good business books written about these topics that some of my more enlightened readers could probably point out, I think the answers really just boil down to common sense. The main thoughts I have focus on incrementalism and modularity.
If your crystal ball is blurry, and can’t see very far in the future, make shorter term plans. Use a more incremental development process. Don’t try to go straight for lunar transportation systems all in one fell swoop. Find intermediate markets that can be served along the way. Find places where you can start growing your future markets using those intermediate ones.
Don’t try to design your whole system as one massive unitary system. Modularity is your friend. Make break points and design in interfaces where other companies can interact more easily with your system as the market develops. The success of open architectures during the PC revolution seem to be particularly interesting. How many of the orginal periphrial companies from back in the 80s still are in business today? Yet we still have plenty of choice with hard drives, disk drives of various sorts, etc. We’ve moved from those big floppy disks to the 3.5″ disks to Zip disks and CD-ROMs, and DVDs and flash drives, all without the whole system having to be redesigned from scratch. How exactly does this apply to space development? I can’t honestly say I’m sure. I think that breaking down transportation systems into modular pieces, with well defined interfaces is a good way to go. It’ll be interesting to see if Boeing or Lockheed eventually sees the light though and decides to become the IBM of space.