It’s fashionable to talk about innovation and failing fast but exploring new technologies has always carried a great deal of risk as a BBC feature on failed aircraft design shows.
Aviation, like automobiles, was a wonderful opportunity for early Twentieth Century tinkerers. With the added impetuous of two world wars, the development of aircraft saw some strange experiments.
One of the things that drove aviation innovation was the evolution of materials science and manufacturing methods, sometimes with tragic results as we saw with the Comet jet liner’s fuselage failures and the DC-10s defective cargo door latches.
In many ways, the early days of airliners was not dissimilar to today’s experiments with smart materials and 3D printing.
Tragedies like the Comet and DC-10 should remind us that in some field the cost of failure is high.When a widget breaks, people can get hurt.
As we experiment with new materials and manufacturing processes, we will make mistakes just as the aviation pioneers did. It’s an ethical aspect of innovation we need to keep in mind, there can be real costs to failing fast.
Image of De Havilland Comet by Clinton Groves through Wikipedia
I am afraid the list of ‘failed’ planes (and other vehicles) is a lot longer . . . .
Let’s start by ignoring military planes, because there may be different considerations applied.
But the list of ‘failed’ planes likley include, for example (super-)constellation, Electra, Tristar (Lokheed).
And strage enough, variants of the ‘failed’ planes flew till recently, or still do (the Comet based Nimrod, the Electra based P3 Orion . . . )
So what went wrong?
The Comet went through pressure cycles faster and more often than previous planes, and the consequences had not been foreseen.
The DC10 was the first plane with a pressure cabin and a (large) door opening to the outside.
Will things still go wrong today?
of course!
– everytime we create a ‘machine’ with new possibilities, we need to understand ALL the issues that it will face and test for it; however, we will likely overlook at least one ‘obvious, simple issue that absolutely cannot create a problem’, but for certain will, as we learn the hard way
– we are not patient enough to do complete and endurance tests, therefore, for quite a few cases we rely on ‘accelerated’ testing: OK when you have a complete understanding of the issues and risks, but not for hiddne problems . . .
Will crowdsourcing increase the reliability and lower the risks of new designs?
It might incite some contributions from specialists, but most likely the crowd will not be familiar with the issues, in particualr the still undiscovered issues, and may in fact contribute to ‘nice and interesting’ solutions that further increase risks.
If I may make a parallel: the Wikipedia is a very usefull repository of information;
but it does not reach the quality of a good enyclopedia . . . .
I do sincerely hope that GE, in spite of their open crowdsourcing, has a team of experts in place that do take the responsibility for the design of the jet engine:
I beleive in democracy for policy choices;
but definitely not for the design of a jet engine . . .
Cees Lanting