Back in the Virtual Saddle

controlProcessFour days this week and then the same next week, I’ll be teaching via laptop and headset. This week it’s hard real-time and next week it’ll be harder real-time – avionics.

The difference is in dealing with the inevitable “Oops!”

What do you do when the computer program that’s been working for ten years suddenly attempts to divide a number by zero? Remember that from High School? It’s a mathematical impossibility. There is no answer. And the computer program is never supposed to do that, but what do you do if some utterly bizarre set of circumstances — maybe an on-board explosion has severed some sensory cables and that now non-existent sensor is being read as “0” — makes it happen.

What do you do then?

If you have souls on-board you might do one thing. That’s next week.

If you don’t have people on-board, then you might do something else. That’s this week.

I was told that a long, long time ago in a far away land, if a cruise missile suddenly forgot where it was supposed to go before going “Bang!”, that it would automatically try to fly to Longitude and Latitude zero, zero. That’s in the Atlantic off the coast of Africa. (Is that a safe place to “drop your load?”)

Regardless, that’s what I was told. That’s what cruise missiles were programmed to do “just in case.”

This week and next I will be teaching two related but different technologies. They differ in how thorough the engineers have to be in dealing with such nearly impossible situations.

One technology predates the other but both have persisted by leap-frogging each other from the 1970s to today. This week (no “souls on board”) is … how can I say it? The consequences we will discuss this week are not as disastrous as those that will be at risk in what next week’s engineers will be building.

Next week’s software flies the 787. Those batteries on the 787 that have now been replaced and re-certified as safe for passenger carrying flights went through the most rigorous of checks before they were certified the first time but everyone who works on avionics knows that you can’t imagine every possibility. It’s not possible. Nor is it easy to figure out what to do. The Air France flight that went down in the southern Atlantic in a storm had the pilot or co-pilot, I forget which, pulling “up” on the stick while the other was pushing “down” — so which one should the computer do?

Those are the kind of things we talk about on break next week and, to a less severe extent, this week.

Fun crowd.

Leave a Reply

Your email address will not be published. Required fields are marked *