One of the things that I find interesting in terms of long lasting standard or infrastructure development is that small variations in design choices early in a system result in radically different performance down the line. For instance the width or guage of a train track will greatly influence the carrying capacity and the size of objects transportable via rail. Same for the size of lane of road. Infrastructure, and standards tend to have an inertia like quality as changeover to a slightly improved system takes significant capital.

In general it seems absolutely insane to me that we rely on an antique control strategy for the Subway while in other industries outside of rail such as aviation or automotive such a system would be easily automated.