Arguably the non-cynical way that the MTA is thinking about it is that if CBTC gives you countdown clocks, bolting on an additional system just to do countdown

Arguably the non-cynical way that the MTA is thinking about it is that if CBTC gives you countdown clocks, bolting on an additional system just to do countdown clocks is extraneous cost. Basically $(CBTC) + $(Separate Countdown Clocks) > $(CBTC). Of course, that's how requirement creep often sets in. Wouldn't it be cheaper to just throw another feature in there while we're at it? That has to be cheaper than building a whole other system right? What is missing in that original function is how adding in the requirement for countdown clocks powered by CBTC increases the cost of CBTC itself. This plays out over dozens of features big and small, and next thing you know you have the F-35. I'm relatively familiar with this story already so I wasn't shocked by the countdown clock portion of the system. I think everyone at the MTA knows that you could build a system that just provided countdown clocks for passengers for not that much money. Keep in mind the ATS system they mention that powers the A division countdown clocks is actually as safety critical as CBTC and therefore is engineered (and priced) like an airliner. It's not priced like an Arduino broadcasting the trains location to receivers that guesstimate when the trains will arrive but only works 95% of the time. The real meat of the story for me is that the MTA doesn't have a realistic plan for rolling out CBTC at all. They basically have the constraint of never shutting down a subway for a significant period of time, and also that to realize the savings from CBTC you have to eliminate the block control system from that stretch of line. It seems like they can't make a hybrid system work unless it covers the entire line (that's why they started with the L). I think if they work on that core problem the whole thing becomes tractable. How do you safely manage the transition of a train from a CBTC "block" to the old style block? If they do that then you can start at the end of the line and incrementally rip out the old system and put in CBTC through weekend work. They were able to replace the entirety of the 7 line's trackage in about 2 summers, and if they didn't have replace the old style signals it sounds like it's not much more work than that to bring in CBTC.