They're funded for a $1-2M study of it on the Green Line. Fallout from the NTSB reports after the fatal D line accident and the non-fatal Government Center wreck. So we'll know in 3 years or so what the official study results say. It's an initiative they'll be strongly compelled to do because of liability risk of another accident and chances that a Strike 3 on operator error accidents will get the NTSB mandating it.
Green probably has most to benefit from it. It's still operating on fully human-controlled 19th century signaling where the inspectors stationed in booths (like the one at the end of the Boylston platform) monitor the schedules on pen and pad. Operators have to obey the very short signal blocks and operate solely on line-of-sight, which is why you have those annoying automatic stop-and-protect signals that are always red like the one between Boylston and Arlington where the tunnel makes a slight descent and the operator can't see the next signal block. Also a reason why they've very reluctant to allow 2 trains into a station at once except where they absolutely have to...they're wholly dependent on the 2nd train's operator to inch in at safe speed.
When you consider that the D is supposed to run all the way through to Medford Hillside the need for automatic signals becomes clear. Especially with the D now speed-restricted after the accident per NTSB orders because of signals that have low visibility in sun glare. CBTC allows for moving blocks, auto-enforced speed restrictions, full computer control from central dispatching to manage headways, and ability to pack the trains much closer at higher speed because the trains' cab signal readouts can sense where the next train ahead is if a block is occupied. No need to operate line-of-sight, all those auto stop-and-protect signals disappear (with possible exception of Copley Jct.), the grade-separated surface branches can operate again at 50 MPH, they can pull into stations back-to-back because an operator would be restricted to 2 or 3 MPH on an occupied block with no potential to cause damage except a jolt and "hard coupling" of the train in front, fewer backups because of fewer red light signals, and if there's a signal failure on the track circuits the onboard computer goes into emergency and enforces an instant stop, then speed-restricted line-of-sight override to crawl over the block. Plus they'll be able to make schedule a LOT more and get actionable computer data to model best way of mixing grade-separated on-signal traffic with the B, C, and E surface routes that still have to obey road signals on full human control like they do today (CBTC would end at the Northeastern, Blandford, and St. Mary's portals just like the subway signals do).
Not at all hard to implement; smarter version of the same tried-and-true ATO system the Red and Orange lines use, with a Positive Train Control layer like the ACSES system Amtrak and the Providence Line use. Catered to non-street light rail use. It's expensive to rip out existing signals and install track circuits...something like $350M for all of the subway and the D. But that price also includes an small order of new cars to fill in the extra schedule slots.
It would be revolutionary. Without a doubt the single biggest improvement they can make to the GL. Even moreso than the surface signal priority Huntington Ave. and Beacon St. are already wired for, but the T won't activate (real surface priority + CBTC in the subway is pretty much the only long-term way to save the B from total asphyxiation).
They've also got an unfunded line item in the 2012-2016 cap improvement budget for study of CBTC on the Blue Line. Less necessary there, but the current trip-arm system of speed enforcement requires switch heaters on the surface at every mechanical signal arm to prevent icing, which gets expensive and failure-prone to maintain with the rough ocean weather. They'd want to eventually upgrade even if the Lynn extension is mothballed just to simplify maintenance. CBTC on a heavy rail line allows for theoretical 2-minute safe headways, again because of the moving blocks and auto-sensing. Far more than Blue has enough cars to handle, but boy...if it ever were extended north to tap the insane ridership projections out to Lynn or Salem...you would be floored by how fast and frequent the service is.
Red is by far the line that needs it most, but the price tag the state PMT specced was $¾B to get all that track and 2 branches done + buying a shitload of new cars to fill the expanded capacity. Orange is almost half bil and may need it least because it's not yet maxing the headways the current signals will support (not enough cars). But that too would benefit if they ever extend on either end or have a need to start using the northside express tracks in full service. Cost on those two lines is why they're angling for Green first (most necessary for safety, most to gain by eliminating human error), Blue second (short, no branches, very few subway stops).
But that would solve literally all of the RL's capacity problems and the downtown choke points. You wouldn't see 2 minute headways because ancient Park and DTX simply do not have the platform sizes or egresses to move that many people off the platform fast enough for dwell times supporting those headways, but it would pad the schedule for the natural variance in dwell times at those old stations so 4-5 minute headways are doable all day without delay. Will never have to stop for a scenic 5 minute Longfellow view because everything's backed up to Broadway. The line would be able to handle thousands more rush hour commuters, which will be necessary with how explosive the load at South Station is growing (to get much worse when the commuter rail/Amtrak tracks are doubled and the waterfront gets built up...to say nothing about whether SL Phase III ever gets built in 30 years). And then stuff like converting the Mattapan line to heavy rail, adding a Braintree infill stop at Neponset, re-studying the Lexington/128 extension, or branching north of Columbia Jct. into the N-S Link are actually supportable in the 50-year range without ripping the shit up out of everything to add express tracks.
Old PMT specs from 2003 on cost (incl. extra cars) and ridership increases:
http://www.bostonmpo.org/bostonmpo/pmt-old/PMT-2.pdf. Dated and only an educated guess because this hasn't been studied at all, but gives you a sense of scale of what the existing track can handle with modern signaling. Blue: +8800 new riders / +2700 all-new transit riders. Red: +9700 new riders / +3400 all-new transit riders. Orange: +10,900 new riders / +4500 all-new transit riders. Doesn't project Green at all because light rail requires much deeper study. $1.3B in 2003 dollars to net 29,400 new daily subway riders and 10,600 all-new daily transit riders who weren't using the system before on 3 lines. At existing stops with not a single extension or single foot of extra track added.
So insanely far and away the biggest ROI the T can get on any system enhancement. This shouldn't even be looked at in terms of an inner city project vs. a suburban investment. All those commuter rail riders have to get around the city somehow; the cascading ridership increases on every mode--bus, CR--make this almost pay for itself. If only the EOT and Beacon Hill weren't so attached to their golden shovels to pretend they're doing something 40 miles out of town they can invest billions they're already lighting on fire with something that'll literally pay itself all the way back by its second decade.
NYC has got this operating on one of its lines as a test. And they've got a much more daunting task with all that MTA track and nothing but old Blue Line-style mechanical signals. But payoff's great enough that even they are looking at making the very painful, very long-term investment.
