My research is related to reliability analysis. I deal with probability and statistics daily. Sometimes I am rather discouraged and sad by people’s opinion on this interesting topic. In my first weeks here, a friend asked me what my research is about. When he heard the word ‘probability’, he spontaneously said, “I don’t like that subject, it’s like gambling”. Mind you, he is also a PhD student; I hope he will understand the importance of probability more than those people working in industry. I replied rather defensively, “Gambling? How come! Probability is a branch of mathematics; you want to say math is like gambling?”.
Another friend, also a Phd student in structural engineering, expressed a similar opinion: “Reliability? such a useless study. I don’t even know what reliability means”. Well, what a pity not to know what reliability means, I thought. You are a civil engineer; surely you know that so many things are uncertain in your structure? Can you say precisely the acceleration of the next-coming earthquake? Or when it will come? Can you say with 100% confidence that your soil cohesion is 25 kPa? Can you be sure that your concrete has a constant modulus of elasticity through the whole beam?
In a prayer-fellowship in YMCA, a friend introduced me to an assistant professor in Civil Eng, NTU. He is also an Indonesian, so we had a good chat for a while. Then he asked me what I’m working on. His response on finding out that I study reliability? “Ah”, he said,”I don’t like that uncertainty issue; I’d rather work with certain things such as concrete behaviour, etc”. Okay, I have a bad news for you, Sir. The only thing that is certain is that nothing is certain (I forgot who said that. But I read that tagline in Vanmarcke’s Random Field book).
One of my uncle is also a structural engineer and he has been working in US for more than 20 years. He also had a similar opinion about my research: “how can you get money from such a theoretical research? Why don’t you study beam-column joint or composite materials instead?”
I’m not an expert in reliability (yet :p) but to me, what I have learnt and understood, however limited, is interesting and beautiful. And the idea is simple. We know that so many things are uncertain, such as the properties of our man-made materials, the frequency and magnitude of a nature-induced loading, etc. A bunch of data of those uncertain things had been diligently recorded. The statistician had provided us with many beautiful distribution, uncertainty quantification, and reliability methods. What can be simpler than applying those theories to represent the uncertainty, and in turn, applying those to our engineering problems? Surely any engineer are more interested in knowing how safe a structure is than merely whether the structure is safe or not?