Tuesday, 12 June 2012

Analyzing Analogies: Prologue

Several years ago, in a different life almost, when I was trying (and ultimately failing) to do innovative and interesting stuff in aerospace, I had an interesting conversation with an older and wiser aerospace entrepreneur.
At one point he remarked that satellites have been getting cheaper because they're like computers, while space launch rockets are like airplanes and that's why they're not getting any cheaper.

I agree with some aspects of this and disagree with others, but the important thing I took from the conversation is not about rockets and spaceships. It's about comparing things.

The question is:
What makes an analogy meaningful?

Before I dive into analogies, I'll inject a bit of TRIZ thinking.

TRIZ acknowledges that a primary hurdle on the road to innovation is psychological inertia.

Psychological inertia is everything you've learned, everything you know or think you know about the problem that helps you do your job and solve routine problems every day, but stops being useful and actually traps you when faced with a problem which requires innovation.
It's a natural part of innovation because innovation is by definition doing things differently than before. You cannot innovate without rejecting existing notions on how things work or should work. TRIZ helps identify and reject these early in the innovation process.

There's another lesson from TRIZ which will serve us here: along it's history TRIZniks have tested many methods of improving innovation, keeping and improving upon those that work and rejecting those that under controlled conditions are shown not to work.

When considering analogies we should keep in mind these two things:
That analogies can represent and contain harmful psychological inertia and that for analogies to survive as a tool for thinking and discussing problem they must be useful.

The Prologue Problem 

Let's look at the Rocket Planes and Satellites Computers analogies.
First, it's an interesting case of using a double analogy to make a point, but I'll avoid commenting on that for now.

The first pair is more interesting since it's these two sets which have not been getting cheaper.
I call them sets because we're not comparing one specific system to another, we're comparing at least a family of systems to another family, or one super system to another. In fact I'll make the point later that what should be compared are something like meta-systems.

So how are rockets and airplanes connected, related or similar?
I'll start by naming the intuitive and obvious reasons:
  1. They both fly.
  2. Made by the same companies and engineers.
  3. They're big. There are small rockets and small planes but they do not immediately come to mind...
  4. They use a lot of fuel.
  5. They're complicated and have a magical or fantastical quality... even to people who understand how they work.
  6. They're both associated with war and military spending.
  7. When they fail they do so catastrophically.
It's easy to see a few of these similarities are problematic.
For example, #2 and #6 are closely related and it's easy to point out that they're at least partially a result of a more fundamental technological similarity. Less obviously, rockets and planes have been strongly associated with each other and with certain organisations because of historical reasons.
It's interesting to note that right after WWI rocket research was more likely to be financed by governments for meteorological research while planes were used early on for postal deliveries.

Reason #1 is also related to #2 and #6 but is far more interesting because it is both very superficial and very significant.

Reason #3 suggests that the sets being compared are often poorly defined and suffer from a selection bias... you might be comparing just the expensive rockets with just the expensive planes!

Can you think of other similarities and connections between rockets and planes? Can you think of good reasons why they should not really be considered similar?

Think about this and come back soon to find out where these thoughts led me...

To be continued...

1 comment:

  1. Disimilar:
    #1: The intended purpose is very different. A Rocket is designed as a one-use object (Exception being the Space Shuttle, but even then, that's not the rocket itself for the most part), while a plane is intended for many, many, many re-uses.

    #2: The manner in which they work is much different. Planes work by the horizontal thrust generating lift via their wings (Or other aptly-named "Lifting Surfaces"), allowing for a much smaller amount of thrust, and thus speed, to be used. Rockets are pure thrust machines. They are moved by the amount of thrust they can produce, which MUST exceed the rocket's mass by a large amount for a significant period of time and distance. Any wings on a rocket are for stability, not lift. On a plane, adding more wings adds more lift - Thats why there were biplanes and triplanes in early aviation, as they needed more lift for their weaker engines. On a rocket, more wings does nothing for the vertical lift.

    #3: Rockets are not used in common, public, commercial endeavors. There aren't many that utilize such technology, as most of those that COULD use it can use aircraft - a much cheaper and reusable item. The only thing rockets truly excel at is getting stuff into space and getting stuff going fast in a short amount of time.

    I think the biggest reason that it's comparable to the planes is #2 - Because there aren't any known methods to truly enhance the capability of a rocket other than make it larger, there isn't really much to research in it. There aren't any new breakthroughs with funky new drives that can be used. The items in the rocket are nearly identical to the items found in rockets 50, 60, or 70 years ago. The only thing we're getting better at is controlling them.