Spacefem (spacefem) wrote,

Engineers, safety, why we're different and why it's complicated

Whenever there's a high profile product safety case in the news, someone asks engineers this question:

"Didn't you want to make your product as SAFE AS POSSIBLE?"

And here's how we respond:

Ghostbusters - When someone asks you if you're a God, you say yes

I just read a wonderful article in The New Yorker about automotive safety (The Engineer’s Lament: Two ways of thinking about automotive safety. By Malcolm Gladwell), and the fundamental difference between engineers and everyone else in the world.

It was about cars - the exploding Ford Pinto, the sudden unintended acceleration of the Toyota fleet, and why these cases were so controversial and sparked so much public outrage. In both cases, people looked dumbfounded at the manufacturers and said "You knew something could be an issue, why didn't you jump to correct it immediately?" And the engineer answer is that we were evaluating whether our fix would definitely work, whether our fix would have saved the last people who were injured in X accident, how likely the issue is to happen again, whether there's a workaround... and everyone hates us for talking about those factors.

I can say this since it's in the New Yorker: there is no such thing as exactly 100% safe. Maybe we can show something like "catastrophic failure condition liklihood less than 10-9" - which, as the article says, makes everybody glaze over then shout "why are you a-holes throwing these numbers at us when we just want you to FIX IT!"

There's a spectrum. Someone could argue that the failure of a coffee maker on our airplanes would deprive the flight crew of caffine, making them sleepy and unable to avoid hazards to the aircraft. And we debate them on that point, so we won't have to install 17 independent power sources to boost the reliability of the coffee maker. We will do exhaustive testing to ensure that the coffee maker won't catch on fire, and our customers will not like the cost of even that testing but we insist, and that's the other half of why nobody likes, or "gets" engineers. We deal with a billion parts, the public hears about two. And on those two we're either too conservative or not conservative enough, every time.

At the end of the day, the only thing I can ever say to convince someone that we are doing the right thing is to remind them that we're all in these vehicles, our families are in these vehicles, we are dedicated to meeting our standards. And the industry leaders who make the standards are also dedicated to setting us up to have safe products, because it's their families in the cars and airplanes. We didn't go into these jobs as a short term money grab. We know people won't buy unsafe products. We want them to be safe.

But when you talk about "what's safe?" That's a question that I'm sorry to say, can't just be simple.
Tags: engineering

Posts from This Journal “engineering” Tag

  • Post a new comment


    Anonymous comments are disabled in this journal

    default userpic

    Your reply will be screened

    Your IP address will be recorded