OK, I’m cheating on this one – I was catching up with Dr John Downer, from the School of Sociology, Politics and International Studies, before he went off to the States for 6 months. But drinking cider on a boat in the evening sun seemed like the perfect time to talk about research (or is that just me?).
Before we build complicated technologies, from nuclear power plants to banking computers, we have to prove that they’re not going to explode or break down in a catastrophic way. However, if we consider a thing with millions of moving parts, working for years on end where all sorts of incredibly unlikely things can happen in that time – proving that it won’t fail is philosophically impossible. But we have to know before we build it, rather than just testing it afterwards. Aircraft are an excellent example of this: it shouldn’t be possible to prove their safety because there are just too many ambiguities involved, but we have lots of aircraft and have accrued billions of hours of flight in total. And it turns out that aircraft about as reliable as we predict them to be, which creates a philosophical problem – given that we shouldn’t be able to prove how safe things are, how come it seems to work? So John is looking at how organisations assess the reliability of aircraft that haven’t been built yet, why that seems to work and the implications for other technologies.
No comments:
Post a Comment