NYC Subway Trains Might Start Moving Faster | Hacker News Reader

NYC Subway Trains Might Start Moving Faster

nytimes.com

80 pointsdaegloe34 comments

I remember reading the article linked in this one, from the Village Voice, at the beginning of this year. Strongly recommend reading it for you armchair MTA critics: it has a lot of tinder for your tirades.

https://www.villagevoice.com/2018/03/13/the-trains-are-slowe...

If my memory serves me well, it made a pretty convincing argument that the MTA was deliberately and publicly misconstruing the "slow train problem" as a result of "congestion", effectively blaming the success of the city for the failure of its public transportation. The Voice article rendered this commonly-touted line of defense not just misleading, but flagrantly incorrect.

From my perspective, it seems that the reversal of policy reported now, only came into existence because of the line of investigative reporting that began with the Village Voice piece.

Credit for your commute's speedup probably belongs, at least in part, to the Voice's fantastic investigative reporting (a full 2 months before the linked NYT piece). Even more reason to support your local journalists! :)

Too late for the Village Voice. It has been effectively shut down. Will there be any real reporting left once Google and Facebook are finished organizing the world's information?

The piece explicitly credits them:

> Workers have also started to replace faulty signals that trigger a train’s emergency brakes at low speeds, a problem investigated by The New York Times and The Village Voice that has also led to slower service.

“Over the summer, Mr. Byford created a new “speed unit” — a three-person team that traveled every mile of track on the system in an empty train to find areas where trains could safely move faster. The team identified 130 locations where the speed limit should be increased. So far, a safety committee at the transit agency has approved 34 locations for speed increases.”

There are almost 100 other possible locations.

> Workers recently started to change speed limit signs on the first segment on the Fourth Avenue line in Brooklyn between 36th Street and 59th Street. Overall, officials plan to change the speed limits at 100 locations by the spring.

The real problem here, as I see it, is that humans are still involved in running the trains. We need speed signs for them to read, we need to hope they follow the rules, we need them to control the train as best a human can. Plenty of cities in the world are running autonomous trains. It's not easy to do, but it is doable. Retrofitting existing trains would be a difficult project, I imagine, but worth the cost.

How much more efficiency could we gain just by shaving those half seconds (or more) at every single start, stop, door open, door close, etc?

I don't even think we need to fire the operators (I'm certain their union would object anyway) so much as change their role to one of managing the autonomous train, providing first aid when needed, keeping a human watch on things.

The problem is the subway has a lot of trouble doing the upgrades that would be required to enable computer driven trains the tracking infrastructure just doesn't exist and installing it requires shutdowns that are extremely disruptive and time consuming.

The MTA will shut down the L train tunnel under the East River for almost two years. There's no talk of automatic trains being introduced as part of that, remarkably long, repair.

Which is incredible since the L line made its first "automatic" debut over a decade ago:

https://www.nydailynews.com/new-york/automated-train-rolls-a...

It costs hundreds of million of dollars and there were service disruptions for years, often entire weekends to do the work. There was no Uber or Lyft either.

The L train has been fully automated for many years.

The reason that only the L was improved was that it is largely separate from the rest of the system. There are extremely few rail connections and none that are regularly used.

Most of the tunnels have 4 tracks, so shutting down even half of them will not disrupt commuting all that much, especially if not done for a very large length. In many places, it will somehow slow down express trains. MTA did this e.g. in Brooklyn on the B/Q line.

I'm afraid that computer-driven trains mostly mean no need for people "driving" the trains, and the trade union won't be happy with that.

> Most of the tunnels have 4 tracks, so shutting down even half of them will not disrupt commuting all that much, especially if not done for a very large length.

I'm going to guess you don't commute on the New York subway. :P

Generally, there's two tracks for each direction, one local and one express. You can't shut down just one direction for obvious reasons.

The New York subway is busy enough that shutting down either the local or express track—while better than taking the subways fully out of commission—is a huge pain. When the express track closes, your commute takes twice as long; when the local track closes, your station will likely be bypassed. And in both cases, this makes for overcrowded trains, to the point where you often have to wait for a couple before you're able to board.

(I'm not saying we shouldn't do upgrades, but shutting down a track is indeed very disruptive.)

MTA's union is heavily opposed to anything that will result in less jobs. Even with the CBTC system that is mostly automated they put two people in the train instead of just one.

I'm pro union and worker's right but the MTA definitely fits lot of the negative stereotypes of unions.

just like literally anything, too much power in some one entity is bad

London's Central Line is automated, however there's still a driver, who presses "go" and opens/closes the doors. More importantly they're also there for if something bad happens or they need to take over.

Is the MTA different?

There are no fully automated train networks even close to the size and complexity of the MTA. Afaik they're mostly really simple point to point trains. Upgrading the 7 line to CBTC costs billions, and that line was expressly chosen because of its limited interactions with other trains.

Making the entire NYC subway "driverless" would probably cost on the order of hundreds of billions of dollars, if it could be done at all. There are more cost effective ways to reduce dwell times, to use your example.

This New Yorker piece on Andy Byford provides a lot more context for why the MTA is the way that it is:

https://www.newyorker.com/magazine/2018/07/09/can-andy-byfor...

Andy did a great job with the Toronto transit. I feel like his work is cut out for him in NYC as MTA looks like a complete shit show from both a political and fiscal standpoint (relative to TTC). I wish him good luck.

My armchair / layperson issue with the subway infrastructure is that a lot of the issues around delays and especially the L train upgrades are related to antiquated signaling systems that seem to not at all acknowledge all of the updates to telecommunications that have been developed in the last 100 years.

I attribute the highly expensive nature of NYC subway to politics, since similar services don't have overall cost anywhere near what New Yorkers pay in fees and taxes. The new chief of the MTA is asking for $40 billion to upgrade the system. They should be able to build three new New York subway systems from scratch with that kind of money.

I would be very interested to see an analysis of exactly why laying track in NYC is more expensive than in other cities/countries. "Politics" is a very nebulous reason—if that is indeed the cause—I'd love to see more detail on where in our political system, and why, etc.

I'm sure part of the problem is that subway workers in NYC are well-paid, but that's probably not all of it.

Fraud, waste, and the worlds worst purchasing work.

NYCT workers are well paid, but that is only an explanation for high operating expense, not obscene capital expenses.

Operating expense is really fairly well managed, albeit with unusually high wages. It is a small, small part of the subway’s problems.

The unusually high wages can also be attributed to a city with an incredibly high cost of living, and large number of overtime hours.

Train conductor base pay is $23-30 an hour - the six-figure salaries that you read about are for people doing >40h/week, including night shift/holiday/on-call work.

I generally think that it's reasonable that people who do more work get paid more then people who do less work... Or that people doing the same work during hardship hours get paid more then people doing their 9-5.

This has been covered in the NY Times [1] and I've read more detailed breakdowns of other projects in more transit/government focused publications. The short version is that the government doesn't do a very good job with planning despite spending a ton on it and the combination of union labor and the hiring structure leads to workers getting paid several hundred dollars an hour during overtime (which happens due to the planning) and more workers performing a task than similar transit efforts elsewhere in the developed world.

[1] https://www.nytimes.com/2017/12/28/nyregion/new-york-subway-...

I assume you're speaking figuratively. Here's an extremely literal elaboration for the railfans in the thread. The route length of the NYC subway system is 245 miles. Subway construction costs vary across developed countries, but generally fall between $200-$500 million per mile[0]. So under extremely optimistic assumptions, 40 billion would get us to about 80% of the current subway network.

For those interested in analysis of subway construction costs, I highly recommend Alon Levy's work at Pedestrian Observations: https://pedestrianobservations.com/construction-costs/

[0]https://www.citylab.com/transportation/2018/01/why-its-so-ex...

> but generally fall between $200-$500 million per mile[0]

Does not help that NYC is paying 5-10x that. Given that, the former comment is pretty much in line, even if slightly exaggerated.

Yeah, I never meant to suggest that it was wrong (except in the most literal sense!). Comparative subway costs do not make MTA look good!

Yeah, but if they hire Oracle or IBM to come up with a solution, as big organizations often do, then you might prefer the 100 year old control system.

Sounds like the MTA is slowing down due to poor maintenance and a lack of active management.

Can't read TFA b/c paywall.

A low maximum speed is not what's responsible for the MTA's deteriorating service levels. The problems are mainly signal issues and train malfunctions[0]. To a lesser extent, track fires. This begets the question -- what good is a marginal top speed increase, if so many trains arent reaching their top speed anyways? At least 30% of the time during my rush-hour MTA commute, my train is crawling through the tunnel, well under its current maximum speed. Think tortoise and hare.

Also -- higher max speed might mean harder stops. [0] https://ny.curbed.com/2018/9/17/17869218/nyc-subway-signal-p...

The change isn't to the top speed, it's to mis-configured time signals that forced trains to slow to a crawl unnecessarily.

In this article (which you probably also won't be able to read, sorry; try incognito?), the Times explains why some mis-configured time signals are the cause of many delays, with compelling evidence to back up the claim.

https://www.nytimes.com/interactive/2018/05/09/nyregion/subw...

Right. Plus how often is an entire line backed up because the operator didn't crawl slowly enough past that misconfigured signal, so the emergency brake got triggered and additional action had to be taken to get the train moving again (release from dispatch? crew sent out?), causing thousands of people systemwide to be "delayed because of a signal problem at 14th St" or whatever?

What if I told you, Neo, that you can read the article, in an incognito/private window?

I've been suspicious that MTA has being keeping the speeds down as a cost savings means to reduce energy consumption. Same with "Sandy repairs" causing service reductions that they won't compensate for on parallel lines.

Their other favorite BS excuse it that delays are caused by people holding doors open. But that doesn't hold water either when uncrowded trains still have the doors malfunction and reopen multiple times.

Pre-Byford management seems to have been controlling their burn rate with these shenanigans. Otherwise the system should be able to handle 1948 ridership levels.

This article explains the slow speeds pretty well: https://www.nytimes.com/interactive/2018/05/09/nyregion/subw...

Basically, it boils down to heavy consequences for operators that go too fast coupled with systems that measure speed inconsistently.

Reply on news.ycombinator.com