The problem with problem solving
Problem solving is at the heart of what drives programmers and many other folk to make things better. The problem is, there may be quite a variety of ways to solve a problem, some of which end up storing up bigger trouble ahead. And VW seems to have fallen into that trap. I would love to know the truth about what went on in the software development teams (or individuals) behind the fiasco, but we’ll probably never really know.
Here is a conversation I’ve made up but which I could quite easily imagine happening:
Manager: We have a problem; we need to reduce the nitrogen oxide emissions to meet US emissions testing standards. Can you fix that? Pronto?
Engineer: Let’s see what I can do.
– some time later –
Manager: Did you fix the emissions problem? Are the cars going to pass the tests?
Engineer: Yes. But we’re going to have to sacrifice fuel efficiency and performance. But those aren’t tested are they?
Manager: No, but that’s what the customer wants when they’re actually driving the things. Isn’t there any way you can get the cars through the tests, and still have good real-world driving performance?
Engineer: Leave it with me…
– some time later –
Engineer: Got it. I used the four-sprung duck technique. We’ll pass the tests now but won’t sacrifice real-world performance and fuel efficiency.
Manager: Excellent! That’s what I like. You saved my bacon. Have a promotion.
There’s a lesson for us all, though.
If the problem is passing the test, then we do whatever it takes to pass the test. In VW’s case that meant making the car behave one way when under official testing and another way when in the real world. Everyone is happy.
But if the test is supposed to be representative of the real world, then we have a massive problem.
You see the engineer solved the wrong problem. He solved the testing problem, not the emissions problem.
To be honest, people are facing the same kinds of pressures the world over and many are solving problems the wrong way, aren’t they? In a school in Uganda in 1998 I was involved with a new head teacher had discovered institutional cheating of exams; teachers would write up the ‘correct’ answers on the board. The problem was student performance in exams and the teachers solved it. Of course the problem wasn’t actually student performance, it was student learning that was the real deficiency. The deficient teaching and its coverup by cheating horrified the new head teacher but solving the real problem took a lot longer.
So here’s something for us all to ponder: am I solving the wrong problems in my life?
Do I cover up my own failures by excuses of weakness or blaming other people? That can work, for a while, with everyone around us. It doesn’t fool God though, who sees right through us. (It probably wouldn’t really fool anyone closer to us.) We truly need an awareness that God knows the truth and that he really can and does and will fix our problems permanently and properly. Jesus dying and cleaning out the noxious emissions of our lives and transforming us fit for a good eternity with him has to be the right solution for the actual problem. And his judgment is final.
ps. The caveat is that we assume in the above that the software was intended to ‘cheat’ the tests. The truth may be more complex. But maybe we’ll never know. Here’s another report about it.