Most bridge engineers will retire having never seen one of their structures experience a ULS event.
We design bridges in Australia using limit state design. A ULS load has a 5% chance of being exceeded over a 100-year life: a 1/2000 chance per year.
A ULS event doesn’t mean collapse, but it does mean plastic behaviour, permanent deformations and likely repairs. A serious event, but not catastrophic.
Now think about your career.
Let’s say you design 30 bridges over 30 years.
With a 1/2000 annual probability, there’s about a 23% chance that one of those bridges has experienced a ULS event.
And a 77% chance that none of them have.
That’s wild - even the most experienced designers in the country have likely never seen one of their structures behave at ULS in the real world.
To me, this reveals two things about our profession.
We’re prone to a false sense of confidence. A structure gets built, looks fine, and we think our judgement was validated. But probability says the structure probably never saw the load case you were arguing about. Whenever someone says “in my experience…” talking about ULS behaviour, it’s probably BS.
Our feedback loop is insanely slow. Software engineers can test, break, iterate and improve in seconds. Bridge engineers might wait decades before the data catches up.
It’s no wonder we skew conservative and cautious with new tech - the system practically forces us to.