Slashdot is powered by your submissions, so send in your scoop

 



Forgot your password?
typodupeerror
Transportation United States

A Look at the NYC Subway's Archaic Signal System 105

New York City's subway system continues to operate largely on analog signal technology installed nearly a century ago, with 85% of the network still relying on mechanical equipment that requires constant human intervention. The outdated system causes approximately 4,000 train delays monthly and represents a technological time capsule in America's largest mass transit system.

Deep inside Brooklyn's Hoyt-Schermerhorn station, transit worker Dyanesha Pryor operates a hulking machine the size of a grand piano by manipulating 24 metal levers that control nearby trains. Each command requires a precise sequence of movements, punctuated by metallic clanking as levers slam into place. When Pryor needs to step away, even for a bathroom break, express service must be rerouted until she returns, forcing all trains onto local tracks.

The antiquated "fixed block" signaling divides tracks into approximately 1,000-foot sections. When a train occupies a block, it cuts off electrical current, providing only a general position rather than precise location data. This imprecision requires maintaining buffer zones between trains, significantly limiting capacity as ridership has grown. Maintenance challenges are also piling up, writes the New York Times. Hundreds of cloth-wrapped wires -- rather than modern rubber insulation -- fill back rooms and are prone to failure. When equipment breaks, replacements often must be custom-made in MTA workshops, as many components have been discontinued for decades.

The Metropolitan Transportation Authority has begun replacing this system with communications-based train control (C.B.T.C.), which uses computers and wireless technology to monitor trains' exact locations. Routes already converted to C.B.T.C., including the L line (2006) and 7 line (2018), consistently show the best on-time performance. However, the $25 million per-mile upgrade program faces uncertain funding after the Trump administration threatened to kill New York's congestion pricing plan, which would provide $3 billion for signal modernization.
This discussion has been archived. No new comments can be posted.

A Look at the NYC Subway's Archaic Signal System

Comments Filter:
  • by Fly Swatter ( 30498 ) on Monday May 05, 2025 @11:41AM (#65353627) Homepage
    The current system may be old, but they can make their own replacement parts. There is no way they can make their own wireless communication equipment when the supplier is no longer available. They will have to upgrade again. And again. And again each time a part becomes unobtainable.
    • I would hope that by this point, with all the subway systems worldwide, that there would be some type of standard, so that manufacturers and suppliers are interchangeable. Or if not, you mandate that part of the contract is to give you the specs to be able to contract someone else to produce an alternative (with maybe a 10 year exclusive agreement to purchase from the original contract winner). There are ways to solve this.

      Plus, the manufacturer would likely be using chips made using off the shelf component

      • There is a standard - ETCS. But it is relatively recent and still quite uncommon.

        • by _merlin ( 160982 )

          Yeah, but there are other, older "standards" with equipment available from multiple vendors. TPWS, ATP, GSM-R, ACSES, PZB, LZB, etc. all have multiple compatible implementations available.

          • And that is the reason why ETCS is uncommon and will stay uncommon - the old stuff is entrenched and still works reasonably well and it is cheaper short term to repair the old stuff than to upgrade.

        • ETCS was created with mainline railways and cross border interoperability in mind. Proprietary CBTC has been doing moving block on metros for years. Practical work on Level 3 ETCS (i.e. the moving block version) seems mostly experiments aimed at reducing the trackside infrastructure for train detection, especially on remote lines. If a standard metro CBTC ever does emerge it seems unlikely to be ETCS based.
      • by ukoda ( 537183 )
        Your background is showing. You are approaching this logically with common sense. The problem is people making choices will have bias from their work history, their influences and incorrectly perceived priorities. This makes them easily influenced by subpar solutions offered by companies that promise to make their life easier.

        It is sad to see but all to common if you don't have smart people at the top with the public good at heart.
    • by kackle ( 910159 ) on Monday May 05, 2025 @11:52AM (#65353651)
      'Good point which I hadn't thought of. I'm tired of hearing from our lessers how "old" is an automatic problem. I wonder whether there's a way of salvaging/improving such a system without going full state-of-the-art, with its IP issues, tariff problems, hack-ability, jam-ability and its lack of transparency as the software elites step in to design and implement their new, complex, opaque system (who will run it after installation?).
      • You could replace pieces of it with more reliable pieces, whether you did or didn't use electronics, but it's probably not realistic to also make upgrades to the functionality at the same time. This system jams itself (literally) which is why they want to upgrade.

    • The question is if this is a new problem, plenty of metro systems out in the world, how do they handle switching and new system and supply chains? Part of choosing suppliers for this is what the expected lifetime and support.

      Ideally the systems are as much domestically sources as possible so even in the case of a future business failure the production can be picked up by someone else as it would be a critical supplier. This isn't a new problem, the military and government already have to deal with it

      • Ideally the systems are as much domestically sources as possible

        Well, they said that they produce/reproduce many of the parts they need in house already....

        Can' t get more domestic than that!!

        Perhaps sticking as much as possible to using the old equipment would be a good plan then...?

        • Huh? That doesn't answer the question of is this a new problem.

          If anything it shows it is in fact not a new problem, at all.

          If you want to keep the old system i want to hear zero out of any politicians or citizens about how much money the MTA needs injected into it every year to keep this old system operating and no more complaints about delays and downtime in America's largest city and 1/3 of the financial hubs of the planet.

    • It appears that the outdated monitoring system is limiting capacity and reducing efficiency.

      If an upgrade (and periodic tech refreshes) are necessary to meet transit demands, then what else can they do? This is far cheaper than digging new tunnels, assuming new tunnels are even possible.

      If they implement a protocol that depends only on basic sensors and simple communication protocols, they could configure replacements cheaply as equipment reaches its end-of-sales or end-of-life age.

    • The current system may be old, but they can make their own replacement parts. There is no way they can make their own wireless communication equipment when the supplier is no longer available. They will have to upgrade again. And again. And again each time a part becomes unobtainable.

      The concept of obsolescence is a bit more complicated than that. Just because you can keep something running doesn't mean you should. Yes the system is currently operational. But is it efficient as can be? Is it as safe as can be? Objectively no for both of those. The signalling system currently used it causing trains to miss their timetables, limits the density of vehicles on the track, and is error prone and dependent on humans. There's material benefit to upgrading the equipment beyond just the thought o

    • by kriston ( 7886 )

      The $25 million per mile just for signalling must be wrong or an exaggeration.

      The DC Metrorail cost about $1-$3 million per mile to build an entirely new railroad from scratch.

      There's a lot of contractor grifting there on the NYC Subway.

    • by AmiMoJo ( 196126 )

      Why can't they make their own wireless communication equipment?

      Look at the Flipper Zero and how many wireless protocols that thing can interact with, using just a single, flexible radio chip (Texas Instruments CC1101).

      These wireless things are not magic, and replacing them with compatible but modern parts is well within the realms of possibility. They should be specifying that the protocol is made available to them anyway, on condition that it only be used once the manufacturer discontinues or the contract

    • They could make their own parts. They just need to redesign out any components that can't be sourced, manufacture the parts and run a training and competency regime to retain enough skilled people to maintain, diagnose, repair and renew the system (this can be very labour intensive for electromechanical interlockings), all to safety critical standards. Since they are amongst the last people using some of that kit, little or none of that cost is amortised over other users and all the risks sit with you. Yo
  • Can anyone tell me why Trump cares at all about congestion pricing in city where he doesn't live?
  • Pretty sure the location issue can be fixed with a fairly cheap solution.

    A bit of python and some Airtags and you've got a map of your trains. You could even spend $30 per train and know where the front and back were.

    --
    When a train goes through a tunnel and it gets dark, you don't throw away the ticket and jump off. You sit still and trust the engineer. - Corrie Ten Boom

    • by Jeremi ( 14640 )

      A bit of python and some Airtags and you've got a map of your trains. You could even spend $30 per train and know where the front and back were.

      Heh. Everybody would know where the train was 5-10 minutes ago. Useful!

      • For this system its very useful. The NY Transit Authority would be 50% less behind with this data.
        --
        You can have data without information, but you cannot have information without data. - Daniel Keys Moran

    • Airtags depend on phones being able to locate themselves which depends on GPS and/or WiFi APs which broadcast SSID. So you forgot the part of your proposal where they install WiFi beacons along the track.

      • by Pascoea ( 968200 )

        So you forgot the part of your proposal where they install WiFi beacons along the track.

        There are probably 5,000 things they forgot in their "proposal". Any time anyone comes up with a solution that involves commodity hardware like air tags, arduinos, or anything similar, it's pretty good guarantee that they've never been anywhere near a legit safety-critical control system. They installed Home Assistant and wired up a few ESP-Home based sensors and that worked, so why not scale it up?

        • I have to admit, and I know you might find this shocking, but I was just being a wiseass.

          I for one know what I don't know. I don't think that because I worked on a narrow gauge railroad and did some Arduinos and shit that I know how to handle subway automation, although I do have some thoughts about it. But mostly they are that I'm glad it's someone else's clearly shitty job.

  • ..AI.

    Not that there's any logical reason for this. But it's the current fad now*. So let's get with it.

    *I could have said blockchain. But that's so last decade.

  • U.S. Threatens to Cut Off M.T.A. Funds Over Subway Safety [nytimes.com]

    “Sean Duffy, the U.S. transportation secretary, demanded a long list of details about crime in the subway and on buses in New York.”
    • Did he get that job because he was on Road Rules back in the 90's? He was very anti-Republican at that time, I see to remember he and the Republican chick having political disagreements at the time. But they did get married, so...
  • by oshkrozz ( 1051896 ) on Monday May 05, 2025 @12:33PM (#65353747)
    Of the pockets $25 million per mile is lining?
    I would like to see an exact accounting of what exactly this $25 million covers.
    I would hope the MTA can learn to negotiate a contract, they don't "have" to upgrade,they would like to, so they would insist that any product they buy (as in purchase, as in own), would come with mandatory right to repair, and full schematics if they need to replace some minor parts. However, the Governor of NY being anti right to repair, and destroying the legislation that was put forward would make sure this doesn't happen.
    • Right to repair does not mean the ability to repair is the most efficient. Objectively we can see now that NYC's system is one of the most repairable there is. It's entirely managed in house. ... and it is crippling performance and hindering improvements. Obsolescence management and the ability to self control equipment isn't the be-all and end-all of procurement choices. There are many variables to consider. In fact in some case obsolescence can even be considered a good thing as it forces an upgrade which

    • I'm guessing $25M a mile includes a lot of the physical infrastructure (the switches and track) along with a digital control system. The NYC subway has a lot of "deferred maintenance" from the last 120 years.
    • by AmiMoJo ( 196126 )

      I don't know but I'd imagine that a lot of it is the cost of shutting down parts of the system to install gear and test it, running two systems in parallel until confidence is reached. Re-training and up-skilling to operate it, and maybe some redundancy money if they are reducing or replacing staff.

      I used to know someone who worked on the British railways, and it was service suspensions and delays that cost the most money.

  • the Trump administration threatened to kill New York's congestion pricing plan

    Yep, NYC needs more pickup trucks on its roads...

    • the Trump administration threatened to kill New York's congestion pricing plan

      Yep, NYC needs more pickup trucks on its roads...

      New Jersey sued New York to block the congestion pricing plan. https://ny1.com/nyc/all-boroug... [ny1.com]

      • by Pascoea ( 968200 )

        New Jersey sued New York to block the congestion pricing plan.

        This single sentence from that article seems to sum up that string of lawsuits: "We’re batting one thousand. We won every lawsuit in every procedural episode so far and we’ll take them as they come,” Lieber said on Friday. “Obviously, the New Jersey lawsuit has been unsuccessful at every stage."

  • Union
  • MTA waste history (Score:5, Interesting)

    by hirschma ( 187820 ) on Monday May 05, 2025 @01:12PM (#65353869)

    NYC resident here. I think that the big issue is that the old mechanical signalling system has been an issue for decades, and that the MTA, even when it was flush, didn't prioritize it.

    Hurricane Sandy took out several lines, and they claimed that they were going to upgrade to digital signalling as part of the restoration process. They didn't do it then.

    The MTA, at one point, had an HR department that made up 25% of it's workforce (!). The history of waste, mismanagement, and misdirected resources is so damning that anything that they do should invite suspicion.

  • by RobinH ( 124750 ) on Monday May 05, 2025 @01:29PM (#65353913) Homepage
    What's wrong with just increasing the ticket price to generate enough revenue to pay for the improvements? Wouldn't those riders be the ones to benefit the most from trains being on-time?
    • Wouldn't those riders be the ones to benefit the most from trains being on-time?

      No, their employers would be, or in the case of corporations, the shareholders. They pocket most of the proceeds from worker productivity.

    • Because that would be like increasing the gas tax until it pays for the roads [archive.org], and we in the USA don't believe in that. We prefer our roads to be socialized, and why should transit be any different?
    • Nothing wrong with a price hike for improvements. The problem is that the MTA had the money before the pandemic, but spent it wastefully.

  • ...easily design a new, reliable system at a reasonable cost
    But only if bureaucrats and profiteers didn't get involved
    What happens instead is that bureaucrats waste a tremendous amount of time and money on paperwork before awarding the contract to a profiteer who sees the contract as a cash cow to be milked for every cent possible while hiring cheap, incompetent people to do the work

    • ...easily design a new, reliable system at a reasonable cost But only if bureaucrats and profiteers didn't get involved What happens instead is that bureaucrats waste a tremendous amount of time and money on paperwork before awarding the contract to a profiteer who sees the contract as a cash cow to be milked for every cent possible while hiring cheap, incompetent people to do the work

      Endless subcontracting is indeed an issue but even if you get rid of that it's still a hard problem. The hard bit is actually doing it whilst keeping the system running. You really can't afford to get this wrong, get it wrong and trains go smack, people die. You can't really close lines for long periods of time without causing huge issues so it needs a lot of careful planning. In many ways it would be simpler to build a set of new lines than upgrades.

  • The US builds a lot of infrastructure then cuts the maintenance budgets. The engineers knew those cloth wires weren't going to last forever, but some politician probably decided that as long as it lasts for one more election cycle, it would become someone else's problem. This happens with schools, nuclear power plants, sewers, etc.

    This isn't about technology obsolescence: NY could replace the cloth-wrapped wires with new cloth-wrapped wires. But they can't just leave them to rot.

  • A far better description would be an electromechanical system.
  • It may be old fashion, but it is immune to malwares. And the concept of buffer between trains seems a good idea to me.
  • I'd like to know how resistant non-mechanical systems will be with regards to flooding. Flooding has gotten a lot more common since I was in NYC. But it occurred from time to time. During hurricane Sandy a good chunk of the system was flooded with saltwater. Something tells me the older systems would fair better than anything modern.
  • New York is the epitome of everything that's wrong with the west.

    It is the epicenter of the occupation and exhibits everything all at once.

    Look at the list of the 100 worst NYC landlords if you want to know who it is. Look it up, it's online and official.

On a paper submitted by a physicist colleague: "This isn't right. This isn't even wrong." -- Wolfgang Pauli

Working...