[TW]Fox;30199283 said:
It's also not that hard to slow down to 12mph either, given that no tram in the UK has crashed since the 1950's, so it was probably quite reasonably thought unnecessary to design such levels of automation into the system. Our heavy rail network doesn't automatically slow down on tight bends either and we've not had a train crash due to excessive speed on a tight bend in living memory!
That's a little inaccurate (Pedant....).
Depending on the curve, speed restriction in force on the curve and other external factors (e.g. risk of incursion onto opposing running lines), there are two main ways whereby which speeds throughout a curve or similar may be regulated.
1. The provision of a permanent AWS (Automatic Warning System) magnet. This would generally be placed at the commencement of safe braking distance from the preceeding line speed, but obviously real-world distances don't always match this.
This will always trigger an in-cab alert, which must be cancelled within 5s, otherwise emergency brakes will be triggered.
Whilst this doesn't control speed per-se, it should provide the necessary jolt to an inattentive driver.
2. The provision of a set of TPWS (Train Protection and Warning System) speed monitoring loops. These get the average speed of the train, and apply emergency brakes if too fast, and would again be placed at a safe braking distance to the curve.
Neither of these will pro-actively monitor the speed within the curve, but aim instead to prevent the train entering the curve in an overspeed situation.
There will also be a degree of feedback to/ from the signaller (Both from lineside equipment and from other railway staff), so that if a train is already running overspeed, or does something unexpected, the signalling infrastructure can be used in order to slow the train appropriately.
Marginally related to this is the fail-safe principle, which is used wherever possible in railway design; For example, if the curve was thought to pose a high-risk of derailments and thus incursion onto other running lines, it would wherever possible be made part of a completely separate section, so that all other trains are held until said section is clear.
TLDR:
There are a whole bunch of different technologies and design considerations working behind the scenes to try and prevent issues like this. Unfortunately these sometimes fail.
The most likely cause IMHO is going to be driver related of some description, which unfortunately it is very hard to design out of the system.
-Leezer-