June 26th, 2011

planet

Primum non nocere

I've always thought about the "how much technology is too much" question, but it has a new perspective now that I've had a baby. And when you're an engineer, the question is incredibly important, because it intersects with your daily life. Women accuse doctors of interfering too much with childbirth. We know that if a carpenter goes into a situation with only a hammer, every problem looks like a nail... that's the problem, some say, with babies being delivered by people who spend a lot of time studying surgery. So I'm in these parenting communities full of smart questioning women, and a lot of them really do not trust anything but homebirth.

Then here I am, an engineer, wanting to make airplanes safer but knowing that all I have is science. I realize airplanes are science, they're man-made machines, but they're flown by humans using our complex brains and muscle memory. Autopilots are great things but even those are complex. The feedback loops try to duplicate our brains. Sometimes they do better than us. Sometimes they don't. In pilot training, I was taught to always question the autopilot, be thinking ahead of it. If it does something you don't like you can disconnect it, if for some odd reason it doesn't disconnect you can overpower it.

My day job involves setting up avionics systems that warn pilots about what's going on with the airplane. That's where the questions come up in my mind now... when am I doing good, and when am I interfering? Every time a warning message pops up on the screen, it's interrupting the pilot. Every time we engineer a message in we'll have to train pilots on what to do, it's more junk to keep in their heads, more pages in the already really big operating manual. And we have to address the question of, "What if it's wrong?" There are some messages that we decide would be "catastrophic", like a traffic warning telling you to CLIMB into traffic when you should DESCEND, so we do these books of heavy analysis to show the probability of failure meets these incredibly low numbers (in my small jet world, 10^-9).

In the world of cockpit designers, one of our more horrific accidents was the 1978 crash of United Flight 173, where 10 people were killed and over 100 injured because the entire flight crew was distracted by a landing gear warning light and didn't pay attention to the fact that they were running out of fuel. They were so busy worrying about that light. Where'd we, the engineers, go wrong? Not enough lights? Too much reliance on one? Definitely not enough emphasis on ignoring technology and flying the airplane.

I trust the establishment. Vaccinate your kid, even if it gives you the jeebies to see all those shots lined up. Buy food from the grocery store. Have your baby at a hospital. Maybe it's because I'm part of an establishment, and I know that even though there's doubt, those people have our interests at heart. I do all the analysis I can to ensure that a NO TAKEOFF message popping up on your display will be good, that it won't come up erroneously while you're rolling off the ground causing you to freak out and abandon takeoff too late and running off the runway. Now that it's on the airplane, am I 100.00000% sure that it's totally necessary? Is the doctor 100.0000% that a woman needs a cesarean section?

On the balance side, I still wanted a natural childbirth, I think it's the healthiest safest thing to do. Going in planning to get an epidural right away with no plan B is, to me, a lot like engaging autopilot the second you're off the runway. Engineers will let you do it, but that doesn't mean it always makes sense.

There has to be a human element. There has to be technology. There has to be a balance.