|
 Talk to any reputable design engineer and they will assure
you of two things:
- Not everything that looks good on paper will work like it
should.
- Multiple design iterations are expected for most projects, especially when entering new territory.
(Credit: PingV Inc. -->)
In any new design or project, we cannot plan to get it right
the first time. And even when we do get it right, there's almost always a way
to improve it or make it better. Herein lies the need for iteration.
Iteration is the act of repeating steps in the design
process (i.e. back to drawing board). After experimentation and testing of a
product prototype, engineers go back to the design phase and make changes to
make it better and/or fix problems. The number of iterations largely depends on
the aims of the design, time constraints,
and resources available for the project. Although iteration is rarely discussed
on its own, it is an essential part of engineering design. Problems with end product designs can often be avoided by increasing the number, quality, and/or speed of iterations.
This is especially the case in the computer industry. Writing
millions of lines of computer code takes a lot of time, and often testing does
too. Bugs in a program can come down to single lines of code, which can make
debugging a nightmare. Thus, each
iteration of evaluation can take a long time to carry out. Increased work time leads to added
costs and/or rushed jobs that result in a lower quality final product.
 In the case of projects which enter completely new
territory, iteration time may be the difference between a success story and
repeated failures. Dr. Paul B. MacCready realized this when he took on the
challenge set by Henry Kremer to create the world's first plane flown by
kinesthetic power. Others who undertook the project spent up to a year
designing and building planes only to watch them fail catastrophically on their
first test runs. To make changes and test again they'd have to reconstruct the
machine all over again. MacCready understood that the bigger problem in this
challenge was not the design difficulty, but the long iteration time.
In other areas of life, long iterations would be completely
unacceptable. Imagine writing an essay in school by hand, and having your first
(and only) draft of it being thrown out by the professor after he or she
pointed out an error. Rewriting the entire essay from scratch after every
mistake would be a complete waste of time! More efficient alternatives would be
to fix the errors on the existing draft, to fix the first draft and recopy it on
a new sheet, or to fix and print changes on a computer file.
(<--Credit: 123RF)
If long rework time is unacceptable in other situations, why would it be a good idea in engineering? MacCready didn't think it was.
He built his plane out of plastic sheet, aluminum tubing, and wire. These lightweight, simple materials reduced the rebuild time of the aircraft from months to
hours. Sometimes he would be able to fly three or four different planes in a
single day because the iterative process (relearn, rebuild, retest) was made much
simpler. Half a year after he started, the Gossamer Condor (shown
right) won the Kremer prize with a successful 2,172 meter flight.
At the end of the day, not all designs will win awards for
success and innovation. But quick iteration will shorten the time it takes to
reach the final design, and that time can often make all the difference.
References
You
Are Solving the Wrong Problem
| 
Re: The Value of Quick Iteration
