Safety modelling
Supervisor: Dr Adam Beagles
What decisions can we safely make from the fact that a system has survived? Typical questions that will be addressed are:
  • Given that n trains have run for m km without incident how confident are we that they can run for another p km?
  • Structures designed with 2.3% probability of failure (BS9708) or even 5% (EN13014) should have multiple failures, but these are not seen. Why not?

Reliability engineering gives a perspective but this project will consider the problem from first principles. The student will consider damage and degradation processes and the stochastic inputs to them. Failure models will be developed for specific systems and simulations undertaken to provide a new perspective on reported risk and failure statistics.