RR503

31 minutes ago, Jemorie said:

Where in my post did I say that the issue with service levels and flagging is running times? I basically implied that the issue with current weekend service levels is flagging. This is where running times have to be adjusted. Look at the printed schedules on some lines for Saturday and Sunday, and you’ll see. In theory, as I stated (before you misinterpreted), you could still increase service. Ruminate. The only difference is speed will be slow instead of normal. That is all. And that is honestly bad service because almost no one who rides the subway plan their trips in accordance to every line’s schedules compared to LIRR, MNRR, and SIR where trains run much less frequently since they are all commuter-based railroads. That’s all I said.

Quite confused. The reason today's service levels exist as they do is because there isn't capacity to run more service because of the signal/flagging interaction I described above. Extended running times are a manifestation of flagging, but have little to do with service levels beyond their linkage to cost/crew availability. 

33 minutes ago, Jemorie said:

And once again, just send every track and signal worker to fully closed lines and call it a day. Because you know that they obviously they cannot do it under traffic. It’s a safety issue compared to no traffic at all whatsoever, which is faster, cheaper, and safer. Also while you’re at it, eliminate the constant sending trains local or express in one or all directions since fully closed lines would be better off anyway. This way, you can increase service on any line you wish. 

I would love full shutdowns, too. They up productivity, create legible service patterns, and force departments to coordinate. But there are some cases where they'll actually reduce the number of trains you can run to a figure below what you'd get by sending all service in a direction local/express/via a different route; not all terminals are created equal. Because of this, and because of the sheer complexity of work planning, there will inevitably be some jobs that have to be done under adjacent track flagging, and I think it's worth considering that there exist ways to eliminate flagging (=its capacity impacts) in those conditions -- that's all I'm saying. 

37 minutes ago, Jemorie said:

But NYC Transit is quite careful when it comes to making sure that service levels meet closer to or within their guidelines. They are not going to increase service just because.

And if you’re really that eager to keep the current extended  after the  shutdown, whatever smh. I just hope you are right that the ridership will increase further. If not, then it stays how it is outside of the  shutdown. Period. I mean you might as well run the  and  on weekends too while you’re at it along with the extended  to and from Flatbush.

Two things:

1) One of the things I've repeatedly objected to through the course of this conversation is NYCT's loading guidelines. They are quite literally myopic. I'd like them to increase service on some lines 'just because' because service increases happens to be one of the best ways of attracting new ridership (though this is contingent on there being ops funding avail for such moves). So the guideline point is absolutely correct relative to current conditions, but, much like adjacent track flagging, should not be seen as an immutable reality. 

2) I'd love it if we ran all those trains on weekends, but I also understand there are real-world trade-offs and constraints. Even in a world without ATF, you'd be hard pressed to fit more than 20-24 tph onto a single track during some "everyone via local" GO operation, so keeping the number of service variants down (ideally to a maximum of 3 services/corridor, which allows each to run at 6-8tph even under GO) is definitely a worthy goal. 

2 hours ago, Jemorie said:

#1 and #4: Fine, you got me yet again....  Doesn't mean you are 100% correct though. Especially your #4 paragraph. You're just making a statement that is more true for the Lefferts Blvd and Far Rockaway branches of the  line, and the whole Rockaway Park  shuttle line, not the rest of the system. The shuttle, in particular, does not need the extra service for obvious reasons. As for the other branches, obviously you know because of Lefferts customers want a one-seat express ride instead of a local first, transfer to express second ride. The only other way I agree with having the  to and from Lefferts Blvd and all  's to and from the Rockaways (mind you, keyword, alternating between Far Rockaway and Rockaway Park, not every single train going to and from one branch whilst the shuttle serving the other) is if the  runs express in Brooklyn, but what will serve the local in return? But that's another topic for another time....anyway, heading back to the  topic:

I should have been clearer: I'm presenting a logical framework here, not saying that this is something that happens *only* on lines that run 3tph. People (myself included) are equally deterred by 5tph service, etc. There's a whole world of transit studies that focuses solely on the way demand responds to changes in headways -- the metric they use is known as headway elasticity of demand. Look it up, it's quite interesting. 

3 hours ago, Jemorie said:

#2: ...and that is where you wonder...why bother disagreeing with me that the  should not be extended outside of the current weekend partial  shutdown when you do not frequent the line? On top of that, it is well over 125% of a seated load because of the . Using your #4 paragraph to prove my point, the  is at 20 minute headways between 8th Avenue and Bedford Avenue, and every 10 minutes the rest of the line. Like you said, nobody wants to plan their schedules accordingly with any line's frequency. This is the subway we're talking about after all. So unless you are referring to the fact that, prior to the current weekend partial  shutdown, the 240-foot long  trains (on 10-minute headways or 12-minute headways) were carrying more than 53 people per car (212 people per train) on 60 foot equipment, I'm not just gonna sweep everything I stated under the rug and take your words to heart. I remember last time when we both had a similar argument about the  (in this case, about shorting some trains at 2nd Avenue) that I should have provided data instead of simply just disagreeing. But you, in particular, have also not provided any data yourself to further prove your point.

Before the  shutdown, the  was a somewhat miserable experience on weekends. Even at its 4 minute headways, it ran well above 125% seated (especially in the evenings). The , for its part, was not at all empty -- it carried a good number of people, but whenever I rode it, it seemed the line's Manhattan connection (to the ) was being used as a 'transfer of last resort,' with many dumping for the  or . So I draw a few objections to cutting service. For one, why? Assuming current work practices stay intact, there's simply nowhere else we can send the train miles -- everyone else runs to the flagging capacity of their line segment. Do we just want to cut service? Returning to the demand issue, though, I would be very interested to see an analysis of the degree to which  converts have been a consequence of the 's absence versus the 's presence (do you see the distinction there?). Giving the  a link to lines that are not the  in Manhattan is quite good for riders -- I would think that a good bit of increased use is more related to that than it is to the 's issues. We'd need a study to do this properly, but given all of the above, I'm quite skeptical that there's no demand. 

3 hours ago, Jemorie said:

If the  on 6th Avenue and/or the  on Queens Blvd are above the 125% seated capacity, whatever, run the extended the . But if it is relatively lightly loaded, then you just run a two or three more trains an hour on the  and  each instead. Other than either of the two options, more service is not warranted. NYC Transit goes by average car load anyway, not by individual trains.

A lil confused here. Post-2021 you *may* be able to squeeze more tph out of the , but the 's interaction with 4th Avenue and Broadway means that it's going to keep getting shot by flagging until NYCT gets its shit together or those corridors get CBTC. 

3 hours ago, Jemorie said:

#3: That is why I said with some exceptions...how did that slip past your head? But you also have to factor that plenty of people already do that on weekdays as well, particularly school field trips, summer campers etc. People who shop on weekends in particular aren't going to take the subway. They would rather just drive instead.

Your original post stated "no one will take the subway unless you don't have a car or unless you don't live near a bus line" relative to people's off peak, intra-boro travel patterns. I merely wished to point out that the exceptions are more like the rule here -- the subway is the *preferred* mode rather than the mode of last resort. To the volume of travel point, yeah, of course there's non-commutation demand on weekdays, but is quite orthogonal to the fact that there's a ton of latent demand on weekends. If you look domestically and internationally, you'll see that some of the greatest achieved ridership increases have been those that upped weekend service levels and took advantage of the relatively elastic demand for transit service during those time periods. Houston got something like a 30% increase in Sunday ridership out of their redesign, for example.

Speaking of totally unsubstantiated assumptions, do provide some data on this supposed preference to drive to shops. As we say on Culver,  is for reight on weekends. 

3 hours ago, Jemorie said:

#5: What more do you want? There's no other way I can think of them increasing service. In order for that to happen, they need to send every worker to fully closed lines instead of concurrently with under traffic. Unless you extend the running times on every line to accommodate the right of way while at the same time increasing service, but that is a an all honestly a pretty terrible option because now trains would run much slower throughout some areas of their routes. Queens Blvd deals with this all the time on weekends as well as every other line concurrently, but only from time to time.

First of all, the issue with service levels and flagging is not running times, it's capacity. Remember, fixed block signal systems are designed to protect a train moving at the maximum attainable speed for an area and thus have long control distances. That's okay if you're running service at normal speeds, but if you reduce it all to 10mph, you end up with a situation where you're enforcing enormous separations between (very) slow moving trains -- that reduces capacity. 

My point, though, is that flagging need not exist in many cases. NYCT actually *has* the beginnings of the necessary equipment to end adjacent track flagging, they just have yet to actually do something with them. I'd start there, and then work towards more shutdown-based GOs (as you say) and some honest to god maintenance productivity reforms. 

3 hours ago, Jova42R said:

How is the fitting on the 's platforms?

3 hours ago, CyclonicTrainLookout said:

The  platforms are wide enough to fit a 5-car set like those on the . It's just that half of the platforms between Court Square and Hoyt-Schermerhorn aren't used since the  are half the length of a typical IND train.

 runs 600' trains (so either 10x60' or 8x75').  line platforms were built for those lengths, they just aren't used fully as @CyclonicTrainLookout says.

1 hour ago, Jemorie said:

Wrong. Subway ridership is down on weekends because alot of people are off from school and work with some exceptions such as public schools having Saturday school classes only for those who failed their regular classes the previous semester or previous marking period. Honestly, how the hell is that comparing chicken and egg?

Is weekend ridership gonna be lower than weekday ridership? Yes. Are current ridership levels indicative of demand? No. My point with the chicken and egg is that current use of weekend trains tells us astonishingly little about how much people want to travel on weekends because the trains, as my friend put it, are always f**ked up. People avoid them, whether that be by Ubering, walking, or simply not travelling. 

1 hour ago, Jemorie said:

Also, I never said anything about reducing service further. I get that you are a highly an informative member of the forums, but nevertheless, the whole reason why I disagree with having the extended  is, once again, ridership. If you are traveling within a borough, no one will take the subway unless you don't have a car or unless you don't live near a bus line. Those who don't work or go to school will instead head out to shop, go to movies, visit friends/family, etc. But that all depends on where exactly they're going in those cases.

Not that I'm a frequent  user, but the times I've ridden it (mostly across the bridge/on lower 6th) it's been well over 125% seated capacity, which tells us that...wait for it...current  service levels are actually less than what is actually required. 

As for ridership patterns, I...do not follow your point. Plenty of people travel to the core on weekends, whether that be because there are plenty of non-work activities in Manhattan or because people frequently work weekends. Intraborough riders also frequently use the subway (when it's running, that is) -- people gravitate towards the system wherever possible, because it's almost always faster than a bus and is (again, when functional) less of a hassle than hailing/owning a car. 

1 hour ago, Jemorie said:

Generally, if trains are not carrying more than the average guideline, then more service isn't needed. In theory, the only way you could do that is if you put any or all track workers and signal workers on completely closed lines instead of both fully closed lines and under traffic.

That is where the problem lies. You get my drift?

Let me give you a scenario. There's a transit service which runs, say, 3 trains per hour on a corridor with a good number of alternatives (a highway, good biking infrastructure and maybe a bus route) and gets relatively little ridership. Is the lack of ridership indicative of no one wanting to use the subway, or is it more a function of no one wanting to have to program 20 minutes of flex time into their schedules because the train only comes 3 times an hour? Extend this logic across the system: are there underutilized subway lines on weekends because there's no demand, or because weekend subway service is, euh, crap? 

As for work, it's not that simple. There exist ways to eliminate adjacent track flagging (track barriers) -- it's just a matter of getting them in use and then convincing service planners at the MTA that they can, in fact, use the newfound weekend capacity to increase service. 

16 minutes ago, Jemorie said:

It’s not needed outside of the partial  shutdown. Subway ridership is different on weekends than on weekdays. If VG8 and VIP are correct that the  is relatively lightly loaded in Manhattan on weekends, even outside of the partial  shutdown, then it isn’t needed.

The average guideline on weekends for 60 foot equipment is 53 people per car and the minimum frequency on any line in the system on weekends is 5 tph (12-minute headways). Once the signals are upgraded, they’ll likely just run more  service than extend the .

Chicken and egg much? Ridership is down because service sucks. The way to change that certainly is not to reduce service more -- though I do agree that in a world where other lines could handle increased weekend frequency, the  is not the highest and best use of crew time. But this isn't that world, sooooo.

Those aren't the guidelines, btw. Minimum frequency on Saturdays is supposed to be a train every 10 minutes, but flagging throws that out the window. I'd love it if every line could run on 6-8 min headways all weekend, but the chance of that happening within my lifetime is, well, quite close to zero. 

3 hours ago, CenSin said:

I wonder how much of that is because they make the late trains skip stops. Then the trains are counted as on-time according to their generous definitions?

If you skip any stop not skipped per schedule, you're counted as late regardless of whether you are or are not at the terminal. 

1 hour ago, CenSin said:

The segment that is out of service is 2 Avenue to Jay Street–MetroTech though. So Bergen Street is within the segment that’s still in service.

That was a miscommunication. The plan had  trains turning north at 2nd Ave and turning south at 4th -- the project was a reboot of Bergen St interlocking. In the end, they just held service for 20 mins or so and turned one  at 2nd Ave.  service just got held. 

2 hours ago, Union Tpke said:

Do you think that any minimal capacity could be gained by using Smith-Ninth to turn trains? Every line should be evaluated for the possibility of full-time closures, and the tradeoffs should be looked at for each option.

Yes, provided that there's staffing at the terminal to double-end trains. I would minimize the use of that one, though. 4th and 7th are super popular stops. You'd probably end up with 7.5tph to 18th Ave,  to Church + maybe 4tph  and then a mix of Smith-9th, Jay (long relay to Bergen lower) and 2nd Ave for the rest. 

9 minutes ago, Union Tpke said:

I think that closing the Culver Line full time south of Church Avenue would have been a great use of a full shutdown. There is CBTC work, station rehabilitation work, interlocking work (Church, 18th, Kings Highway), and now ADA work (Neptune, Avenue I). This was one of the corridors that was proposed for overnight shuttle bus service in the 1990s.

Absolutely, and in the case of Culver they knew about everything but the ADA initiative well in advance. The issues with Culver shutdown, though, are more real than with those other two lines. You have terminal capacity and yard access problems. On the former issue, even if you did double end every train and eliminate train clearing, you'd still get the +25 seconds from DGTs, along with speed penalties from the 3-yellow move to the yard, the GT10s on the yard ramp, and the AKs in the yard itself. Especially since those issues get worse with longer trains, I dunno if you could do full  service out of that terminal.

As for yard access, unlike Dyre and Sea Beach where yards are either not on the line or are accessible via alternate routes, there's no way for s to access CI if lower Culver is shut. You could fill the express tracks north of Church with layups like they do for the cold weather plan, but that's not great for crews and may not provide adequate layup capacity. There's also the issue of you having to keep lower Culver open to at least 18th to get trains out of those tracks and onto the local, unless you wanna come south from 4th Ave. 

These issues are likely tractable--you could keep 18th Ave open as an overflow terminal and hope that takes enough of a load off of Church (if not, 2nd Ave), and you could just bite the bullet on layups--but they're issues nonetheless.

3 minutes ago, Union Tpke said:

The same should have been done for Dyre Avenue, which in recent years has had its signal system replaced, structural work (Dyre), station rehabs (Gun Hill Road), track work, and maybe other work I have not been aware of. They have undergone too many years of nonstop GOs.

Agree, though the grid up there isn't all that conducive to efficient shuttle replacement service. The treatment of the Astoria line has likewise been shameful -- switch replacement projects, station renewals, ADA Astoria Boulevard, track panel replacement and now CBTC happening essentially without GO coordination. Some of this, to be fair, was a consequence of opacity as to future initiatives, but that's a controllable flaw -- a major project on a line should trigger conversations about potential concurrent initiatives. 

33 minutes ago, Around the Horn said:

Sea Beach only had two active tracks during the rehab period and the closest interlocking is further down the line at Kings Highway. That long relay would have killed reliability even more than the express in one direction, local in the other actually did.

Not to mention that the main transfer point that they're trying to get everyone to use for this round of work was completely closed to install elevators and trains on the express track skipped it anyway which really makes this idea a non starter.

The proximity of Sea Beach to its alternatives, the insane, access-driven duration of the station work, and the fact that travel patterns were being seriously disrupted anyway would suggest that a full shutdown may have been a net positive here. s via West End (or to Bay Ridge), shuttle buses, increased  service, etc and just be done with it ASAP and all at once. 

Would have likely saved some serious $$$ too. 

You’ve gotta respect the lengths to which this city will go to avoid uncomplicatedly adopting international best practices. Like seriously folks, CBTC systems that support 40+tph provided halfway decent ops are available off the shelf these days. It’s not that hard. This country’s fascination with ~~~innovation~~~ can be quite loony at times, and the institutional effects of systematically ignoring what everyone is telling you to do are great. At any rate, isn’t reducing customized unicorn products one of the goals they identified in the capital plan? 

There’s a joke to be made here about “how many academics does it take to install a signal system” but it is alas out of my grasp. 

55 minutes ago, Deucey said:

Solid state interlocking?

Interlocking is run by computers rather than by electromechanical relays. 

10 minutes ago, CenSin said:

That interlocking didn’t get any work during the Culver rehab 10 years ago?

No, it was NYCT's first solid state interlocking. Cut in in 2004 or thereabouts. 'Twas a disaster -- the thing is a lemon, and its (poorly signed, questionably calibrated) sea of GTs easily lose 90 seconds against what existed previously. I'm thankful it's being replaced, though am a bit apprehensive we may get something even more wacky. 

Full report is out

https://new.mta.info/sites/default/files/2019-09/MTA 2020-2024 Capital Program - Full Report.pdf#page=14

I would post highlights, but there's honestly nothing too earth shattering in here. They're replacing Bergen St interlocking, which is funny, and it seems we may get work train CBTC, but beyond that the stuff in here is either super unspecific or already in the public domain. 

1 hour ago, Around the Horn said:

Could we do this at Woodhaven Blvd on the  though?

Unsure. One of the great advantages of Westchester Sq is that the station isn't abutted by buildings on its north side. Not the case at Woodhaven, though I'm sure there's some way to phase things to swing a conversion. (build out a length of middle track around the station and have local trains bypass in one direction while their side is rebuilt??) 

3 hours ago, LaGuardia Link N Tra said:

I’m a little confused at this part right here. What is this track map supposed to represent and could it work for stations like Rockaway Blvd on the ? Heck, could a similar method even be used to fix the design flaw at 72nd Street/2nd Avenue?

See the quoted post -- this is a conceptual phasing to convert Westchester Square  to an express stop so that local trains can turn there rather than at Parkchester.

It could work at Rockaway Blvd, though I don't know why you'd want to do that, and it'd be impossible at 72 without *major* spending. 

1 hour ago, Union Tpke said:

@RR503 This is incredible. $4.7 BILLION will be in the Capital Program for Penn Station Access, in addition to the $895 Million in the 2015-19 Program. The whole advantage of the project is that it is using EXISTING rights-of-way, making use of EXISTING tracks. What the heck went wrong? I just can't..... All that is involved is some track work, substations, a bit of electrification, and new stations. UGH!

 

https://www.lohud.com/story/news/2019/09/16/mtas-51-5-b-capital-plan-help-pay-penn-station-access/2341595001/

This is giving me heartburn. Also reinforces my thesis that the future of American urbanism is on the West Coast. NYC is so stuck, and without really anyone who can unstick it meaningfully. 

Short turn trains do not address the fundamental limitations of an interlined upper SAS, whether you see those as the unreliability caused by the merge at 72, the negative capacity impact on Broadway service, or the fact that lower SAS would never be able to reach its full potential. If I had to give just one suggestion, it would be to rework SAS. I and others have proposed a whole number of schemes to do so. I'd be happy to rehash those if it'd be helpful. 

The "structure may not support it" argument feels like quite the deus ex machina in the context of a plan as extensive as this. We can build subways out into Queens, a 4-track SAS and a new East River tunnel, but hell is gonna freeze over if we make some ironwork mods to an elevated structure at one of the widest points in its ROW...? Methinks that this is possible, and indeed advisable. I'm making this up in between research tasks so don't shoot me if there's some super obvious flaw, but it would seem like you could do either of the following: 

Is this the best way to do it? Probably not. But I daresay this is a relatively tractable issue. 

Now, to the points about ridership. Co-Op City has high bus ridership in part because it has no subway service. Untold thousands board buses every morning to get to PBP, stops on Dyre, etc. The number of people riding buses in the area would decrease dramatically with the number of subway extensions you're attaching to the area, and given that Co-Op is relatively difficult to access from the surrounding street grid and that areas even just over the highway from Co-Op have good access to the , I would not expect it to become a bus transfer hub as is Flushing. This is all to say that running express service all the way up there is a bad idea. You've correctly identified why the  works, but the  doesn't have those characteristics. Serving some local stops before commencing the express segment is the move here, and will help keep car equipment requirements and operating costs down in the long run. 

On 9/16/2019 at 3:29 PM, JeremiahC99 said:

I started thinking of a South 4th Street Subway because in addition to changing the  from serving 6th Avenue to 8th Avenue, I was proposing a few other changes I did not include. For one, I was thinking of consolidation of stations on the Jamaica and Myrtle Avenue Elevateds, rebuilding the Myrtle Junction, etc. Those plans, combined with one of your suggestions of all  and  service over the bridge made me think "If I want to do all of this, I might as well make a whole new subway line while I'm at it. We have provisions for a Worth Street subway and that can feed into the new line." Form there, I would do the vanshnookenraggen plan and have 8th Avenue service serve thre Worth Street Line and reroute 6th Avenue express service to Williamsburg. More specifically, I am proposing a slightly-modified version of this: 

You certainly _could_ do this, I just don't know whether the value equation works out. Spending precious dollars on 1:1 replacement of elevated infrastructure is very IND, and is not a luxury we have today. I think you're much better off with the incremental improvements we've discussed--they're unglamorous, but they're also cheap, and crucially do not involve building even more underriver tunnels. 

I do want you to do that, and I very much stand by that want. Routing  to the local tracks of Fulton will serve to reduce the capacity of their Manhattan trunk segments through reverse branching and will just be an annoying source of delay. You really should reconsider this. Generally speaking, I think we all need to step back and reconsider the ways we're integrating SAS into our subway plans. The best ideas out there IMO are ones which combine segregation on the north end (ie no reverse branch at 72) with proper integration at south (SAS-Nassau, SAS-Manhattan Bridge, etc). I think that's the way to go. 

I detailed the CBTC-merge capacity issue in another thread, which I hope helps you understand why those 40tph capacity numbers are somewhat dependent on their not being too many merges. For these reasons, as well as those of reliability, I think deinterlining QB is the way to go here.  local,  express/ extension, you take your pick. 

9 hours ago, Calvin said:

With this short-turn at 86 St on the  next month, in November, will they increase the headways on the Coney Island side instead of waiting 12-15 mins on the weekends to the regular 8-10 minutes?

You can’t reduce weekend  headways. Too much flagging. 

9 minutes ago, P3F said:

@RR503 Any idea of the potential terminal capacity of the 86th Street relay? I haven't checked the schedule in a while, but I believe the runs a max of 10 tph including the ones signed as via Sea Beach.

Yes, it's 10. It'll actually be really interesting to see what happens. I haven't ever observed relays at 86, so am unsure of how exactly they run it. The constraint I see is the fact that the x-over just south of the station is only in the facing-point direction, which will likely limit them to a one-pocket op. You can turn 10tph on a single pocket, but that does require some basic discipline that is sometimes lacking at NYCT.

2 hours ago, Jsunflyguy said:

There is only one supplement for  since 8/30 that I can see.

So the gap remains, then?

2 hours ago, CenSin said:

And that couldn’t be reclassified as some other kind of “stop” without too much red tape/programming difficulty? Is the input data totally unavailable for the computer to make that kind of decision?

I don't know. You'd basically be telling the computer that it's okay to overrun a stop arm in cases where you're facing a home red for lineup and not occupancy reasons. Unsure that'd fly with system safety, though I do know they made some change to Canarise CBTC interlocking logic that had something to do with speed...

3 hours ago, CenSin said:

Hmm! I thought CBTC would actually do merges better by allowing trains to be closer together where there are interlockings. I’ve seen when trains cross over just past the end of the platform and the next inbound train on a different track is held outside the station even though there would be no physical interference. I’ve also seen  trains during AM rush come into 36 Street on the local track just to facilitate a more customer-friendly merge north of the station instead of holding one train in the tunnel. CBTC ought to be able to make decisions like that on-the-fly.

Let me illustrate my point using Canal St. A southbound  entering there usually gets a short route into the station, but faces a red home signal at the leaving end of the platform. That red home does not indicate a conflict so much as it does a lack of an established route all the way through the interlocking; a train tripped by it would likely overshoot onto the local track without any issue. Aside from the weird approach locking time thing that happens north of the station, operators enter in these conditions as if everything were normal -- they may brake a bit more conservatively because they're facing a red, but they do not act as if they're entering a stub-end terminal. CBTC, however, would enforce that level of safety. If my understanding of NYCT CBTC architecture is correct, a stop arm is considered a fixed obstruction, and therefore becomes the endpoint for a safe braking curve which, in turn, forces trains to enter interlocking areas at restricted speed -- imagine the  entering 8th Ave, or the  entering Main. 

On interlocking CBTC islands, I don’t think that’d work well. CBTC’s most positive effects on line capacity are at crowded stations — you sometimes actually lose capacity around merges because CBTC treats stop arms like they’re bumping blocks. In any case, many of the merge fluidity benefits of CBTC are accrued through reduced runtime variability over the course of the route (=better OTP at merges), which is a benefit you’d lose here.