Keep It Simple Stupid

The 248-page Columbia accident report came out in August 2003, six months after the spacecraft disintegrated during reentry. Seven astronauts perished on February 1, 2003.

I read the report the day it was released. When I got to page 14, I took out my red pen and underlined the following statement: "The Shuttle is one of the most complex machines ever devised." At the top of the page I spelled out in capital letters: "KISS" and added "That's the problem!"

I first heard of the KISS principle—Keep It Simple, Stupid—in 1979 when I started working at JPL. I found it puzzling. I never met anyone at the lab that I thought was even remotely stupid. Twenty percent of the employees had Ph.D.s.—nearly a thousand doctors of virtually every branch of science and engineering. There were aerospace, electrical and mechanical engineers, and astronomers, astrophysicists, and mathematicians. Every once in a while I ran into a less obvious specialist: a biologist, a philosopher, a soil mechanic.

So who were they referring to as "Stupid?" (A Navy guy told me I had it wrong—it was supposed to be "Keep It Simple, Sailor," but I am discarding this hypothesis.) The people at the lab weren't stupid. Then I heard Richard Feynman, who often gave lectures to the Caltech-JPL community, say (apparently referring to himself), "Even Nobel laureates can say and do stupid things."

What JPLers were doing was anything but simple. Dual-spin dynamics, multi-body celestial mechanics, tests of general relativity, and deep-space navigation are by their very natures exceedingly complex subjects involving advanced mathematics, millions of lines of computer code, and billion-dollar spacecraft. How can you keep it simple?

The idea (and this took me a while to get) was not to make it any more complicated than it already was. Albert Einstein put it this way: "Everything should be made as simple as possible—but no simpler."

There is a tendency for rocket scientists to sound like rocket scientists instead of human beings, analogous to students trying to sound "literary" when asked by their English teacher to write an essay. These bad writing habits seem to originate in grade school with the assignment, "How I spent my summer vacation," where the kids tell the teacher what they think she wants to hear.

The neophyte rocket scientist tries to impress his boss by demonstrating his Ph.D. prowess: gobs of equations, recondite theorems, reams of numbers. The new rocket scientist thinks his boss will appreciate this morass of technical detail. Just like the essay student who tries to please her teacher with big words and complex sentences—and without a clear message or story.

So the boss issues a caveat to all new Ph.D.s: follow the KISS principle.

There is more to this concept, however. It is a well-known principle of design that simpler systems have fewer failure modes. Simpler systems provide the anti-Murphy strategy: If there is less that can go wrong, then less will.

A simple capsule design got men to the moon and back. These capsules were robust, easy to fly, and had simple backup systems. Now consider the shuttle: It has millions of parts, it is difficult to fly, and it has few viable backup systems. It is the very antithesis of simplicity.

The KISS principle was completely ignored when NASA designed and built the shuttle.

0 0

Post a comment