Beyond the route: Introducing granular MTA bus speed data | Hacker News Reader

Beyond the route: Introducing granular MTA bus speed data

new.mta.info

217 pointsNelkins73 comments

A fun fact about NYC's buses: many of the routes are turn-by-turn replicas of previous streetcar routes; Brooklyn alone had dozens[1]. The B46[2], for example, follows the Utica-Reid line as it ran until 1951.

They never actually tore up most of these lines; the city just paved over them. You can see them poking through the pavement whenever the city redoes the roads.

[1]: https://en.wikipedia.org/wiki/List_of_streetcar_lines_in_Bro...

[2]: https://en.wikipedia.org/wiki/B46_(New_York_City_bus)

Ignoring the huge issue of political will, how much more or less effective would street cars be on these lines, compared to buses?

I think it would depend: one of the reasons the streetcars were originally eliminated is that they were increasingly held up in traffic, and the argument was that buses could navigate (like cars) around traffic, make detours, etc.

In practice however that hasn't really been borne out: the city's buses are notoriously slow. The city has (correctly) reprioritized bus lanes (including lane enforcement for scofflaw drivers) and express services (SBS) in response, but at that point we're essentially back to rights-of-way (i.e. how much of Europe runs timely and efficient streetcar networks).

In short: I think streetcars would be less effective if not (partially) separated, but more effective otherwise. Given that the city is moving towards bus lane separation anyways, I personally believe they should revitalize the streetcar network instead. But that's (1) expensive, and (2) involves impressive amounts of local political spaghetti, given that the buses are currently run by state-level MTA while the roads are owned by the city.

Ya the idea that buses are better than streetcars because they can go around traffic is just completely detached from reality. Maybe a bus can go around one double-parked car but during rush hour that's not happening. It was always just an excuse to avoid taking an inch of space away from cars for dedicated transit right of way.

I live in Toronto and avoid taking streetcars because I have seen significant delays precisely because of parked cars or some other blockage.

Of course in my opinion this simply means banning parking on streetcar routes; or having dedicated streetcar lanes where a blockage simply isn’t possible.

Streetcar routes in Toronto which do not share the road are fine.

What's the benefit to streetcars over busses with a dedicated, physically separated right of way?

I like the idea of streetcars, but busses seem easier to purchase than streetcars, standard road paving seems easier to maintain than streetcar tracks and power, and likely it's easier to find/train bus operators than streetcar operators (even though I assume streetcars are actually a bit easier to operate).

The main one, in my mind, is permanency: as I mentioned in the adjacent comment, stable car-independent communities tend to be built around transportation systems that can’t be easily removed.

(I think there are other benefits, like being slightly more comfortable. But permanency is by far the most important.)

Capacity and level boarding are the two big ones.

Trams are more capacious than buses because they don’t have onboard fuel tanks, so more space for passengers; and they’re fixed to tracks so they can be significantly longer without worrying about the back swinging out.

Trams are also perfectly level with platforms, so there’s no need to waste time to achieve level boarding for wheelchairs, strollers, luggage and the elderly; buses can spend quite a lot of time kneeling and deploying ramps.

These are good points!

In prague there is both extensive bus network and tram network. I almost always go for buses. The capacity is just so much higher and usually the drive is much smoother compared to buses. Also trams are powered by electricity making it more efficient and c02 neutral...

> standard road paving seems easier to maintain than streetcar tracks

I would think that tracks last way longer.

Overall I think the cost is lower in long term for street cars but the initial cost is super high - e.g. edinburgh build one awkward tram line for around 700m. But thats with depots, cars everything. In Prague with all existing infrastructure it cost now about 78m usd to build 2.2km of tram with 6 stops.

Fellow European here. My understanding is street cars started out as futuristic marvels of modernity, but unlike their cousins trains & subways, they aged fairly poorly and don't generally do well in mixed city traffic today:

First, you can’t go faster than cars or avoid traffic (in practice), so there’s no obvious advantage like with trains. Secondly, buses got a lot cleaner, spacious, comfortable and quieter. The modern buses in European cities are not just on-par, but often more comfortable and allow higher speed on long stretches, because modern suspension beats aging fixed rail (it tends to be shaky, again unlike trains). So then what’s the point? Trams are electric? Given how buses are basically commodity in our oil-centric world, I can only imagine how trams look at the balance sheet in comparison.

Now, there are some exceptional cases where I really like trams. When the route has majority separate rail (typically in beautiful stretches of nature) but can switch into streets when needed to reach better. For instance, Tvärbanan in Stockholm is a tram that – while not always perfect – is universally appreciated by most.

> universally appreciated by most

60% of the time, it works every time

I really like the idea that street cars, trains and subways could share a single network (kinda like they do in Tokyo, except Tokyo doesn't really have street cars, mostly trains - https://www.youtube.com/watch?v=0KMYAEIXVzA).

It would allow trains to come from one direction, pass through the city undisturbed and emerge on the other side and continue.

Known as tram-trains, and an established model in Karlsruhe and Kassel, Germany.

The federal rail administration would never let this happen in the US

> First, you can’t go faster than cars or avoid traffic (in practice),

that is not true, in cities the car speed is usually limited to 50, a lot of trams go 70 on certain sections. Also "or avoid traffic" a lot of trams go completely separetely from the traffic.

> because modern suspension beats aging fixed rail (it tends to be shaky, again unlike trains).

Depends on the city, but a lot of cities that I visited have a very modern trams that are not shaky (helsinky, zurich, bratislava, riga, edinburgh, bordeaux...). Also the technology of the rail building has changed and the new lines are meant to be quiter and more stable

> Trams are electric? Given how buses are basically commodity in our oil-centric world, I can only imagine how trams look at the balance sheet in comparison

No idea what you mean by this but I would assume that the cost of running things is lower, the c02 profile is for sure https://ourworldindata.org/travel-carbon-footprint

> No idea what you mean by this but I would assume that the cost of running things is lower

I meant that light rail must be much more expensive, but now I’m not so sure. I hadn’t considered you can have more passengers per driver and if labor is dominating cost then yeah trams can be cheaper!

> a lot of cities that I visited […]

Have you accounted for the reliability of these networks? In my experience trams (or rather tracks and electrical- and signal systems) often break down when there’s snow in the winter, leaves in the fall or sun-bending in the summer, which may not be noticeable on visits. That can also increase costs, since the backup is usually buses and you need a task force who can go fix problems.

Maybe I’ve been unlucky, but my experiences relying on them everyday (in San Francisco and Gothenburg) have been disappointing.. it feels like those networks have been kept alive for nostalgic reasons.

> I meant that light rail must be much more expensive, but now I’m not so sure. I hadn’t considered you can have more passengers per driver and if labor is dominating cost then yeah trams can be cheaper!

Well labor and gas/electricity. I think in most places in EU the electricity will win over gas easily.

> often break down when there’s snow in the winter, leaves in the fall or sun-bending in the summer, which may not be noticeable on visits.

Not an issue in Prague where I have experienced them the most. You can clean the tracks in similar way that you would clean the road, so the buses would not have much of an advantage..

> it feels like those networks have been kept alive for nostalgic reasons.

i have not been there, so maybe its true tho Gothenburg seems to be investing into the network and buying new trams. So are a lot of other cities.

there are a few

* recently a big trend is grass tramways. generally speaking this is more ecologically friendly by reducing impervious surfaces and replacing it with greenery, which generally lowers the urban heat island effect and is better for stormwater absorption. as a nice side effect, it is also generally a more visible differentiator from car lanes that people are less willing to drive over.

* trams are generally more capacious than buses because they are laid out better for more standing room. they are also more capacious because it is safer to run very long trams since the tram is fixed to the tracks; there are practical limits to how long a bus can be since a driver needs to be careful when switching lanes and whatnot. The longest single tram unit is 58m, the longest single bus is 32m; and you can couple trams together.

* trams don't really move side to side due to being fixed to tracks, so level boarding with little to no gap is much more realistic to achieve than on buses. This is generally much better for accessibility and speeds up boarding time; if you've spent any time riding a city bus, even a low floor bus spends a significant amount of time kneeling to achieve worse results for level boarding. And buses kneel not only for people in wheelchairs, but for people with strollers, with luggage, the elderly, etc.

Would just like to note one issue i have observed with the MPLS light rail: multi-car transit has less oversight and is more attractive for drug use and shelter for the homeless which lowers use by commuters. Our busses running the same routes are safer and better options.

I haven't seen these problems on the Minneapolis Metro even riding at night, but if it is actually a problem, it seems like the solution is build out actual infrastructure to support the homeless community.

Which Minneapolis very much does not have right now despite the best efforts of one or two plucky underfunded nonprofits.

If your light rail cars are the best option people have, that's not an issue with the transit design, that's an issue with the rest of the infrastructure

It’s a severe problem in mpls. Are you sure we’re talking about the same city and rail system? Never heard it called “the metro”. I was born and raised in DC where that’s what folks call the subway… only ever heard it called the light rail…

The rapid transit system in Minneapolis, MN is called the Metro. Technically that includes the bus rapid transit system, but I'm just talking about the light rail since I do prefer it over the buses.

All I can say is I've never seen any issues on the LRT. I'm genuinely sorry you have.

Streetcars can be up to 100m (300ft) long and fit 1000 people, saving costs for drivers.

Roads works are expensive [citation needed], and buses are heavy [citation needed], so they cause lots of damage over time. These costs are often not tracked correctly. Metallic rails require far less maintenance, but have a higher initial cost.

Trams can be quieter than ICE buses.

If build appropriately, Teams can even take sharper turns than buses.

Any public transport that doesn't have its own right of way is immediately and trivially broken, since it will be always strictly worse than an individual car. So more people will take cars, so traffic will be worse, so buses will be worse by the same measure, so will still be worse than cars, so... That's how you get LA levels of gridlock despite every street being a 14 lane freeway.

Whereas if buses/trams run in a dedicated lane with the same speed independent of car traffic, there is immediately a natural balancing incentive / restoring force: too much car traffic and the bus will become comparatively more attractive, so less people will take cars, so traffic will be lessened, so

NYC buses have been crowded since before I was born 30+ years ago, even if they are crawling at 4 mph

> the city's buses are notoriously slow

Are the buses actually slower than they can be, or do they just have to deal with too many passengers (given the population density) taking a long time to board/unboard?

Boarding time is a factor, but traffic, road conditions (double parking and driving in the bus lane), and stop spacing are larger ones. Especially since OMNY and pre-purchased tickets for the SBS routes.

Enforcing right of way for street cars could be a lot more efficient now. Stick a licence plate reader on the front of trolley, record any car that blocks them for more than X minutes and mail a ticket. Drivers will learn fast not to screw around.

> for more than X minutes

You meant seconds there, not minutes.

For two-lanes-or-less-each-way streets, buses don't offer much benefit if the streets are busy since the chance of being able to get enough space for a very heavy, long bus to switch lanes is low/impossible.

Streets cars/trams weakness include breakdowns - trailing streetcars/trams are blocked if a streetcar breaks down on the track. But, they can also carry more passengers than buses. I think many modern versions are about 3-4 bus-lengths long.

From the perspective of "vehicles on the road" buses make a lot more sense to me:

- They can maneuver around double-parked cars and trucks

- They can switch up the route when there's construction

- There are no tracks tripping up pedestrians and cyclists

- They're [probably] easier to get to a service hub for maintenance

- They don't require overhead wires to provide electricity

- I would guess they're cheaper to purchase and maintain, but don't have a reference

One area where street cars _might_ win is noise. Busses can be loud.

Your last four points are good, but in practice the first two have not netted significant advantages for NYC's bus operations: many of NYC's buses run on narrow one-lane streets, where any amount of double parking makes the road completely un-navigable. Similarly, it's more common to see a bus route taken out of operation entirely for a week than to have it re-routed on the fly (the latter does happen, but the network also dense enough where most riders can take the next avenue's route).

I think a significant understated advantage to streetcars is their effect on local neighborhood development: like a subway line, a streetcar line is a semi-permanent installation that can't be easily taken away by a short-term replanning of the network. Bus lines, even when dense and well-developed (like NYC's are!), simply feel impermanent in a way that rail transport doesn't.

(Or as another framing: if you build a rail connection to a neighborhood, there's a good chance there will still be a thriving neighborhood there in a century. It's not as easy to guarantee that with a bus route that can be taken away overnight.)

Rail-based transit also provides major side-benefits to its routes: development and improvement.

The principle is that bus routes can change, bus stops can move. Rail right-of-way and train stations are quite permanent and immobile.

Therefore, if a city invests in rail, the developers will follow, and redevelop, revitalize, or gentrify neighborhoods along that route. Conversely, folks in the neighborhood may fight the rail expansion, because "there goes the neighborhood" usually in a more upscale fashion.

It was smart for cities to build out streetcar lines in their early expansions, enticing developers into areas that promised long-term access. Of course, rail lines don't last forever, but the point is being more permanent and staying put, more reliably, than rubber-tire-based transit.

Light rail provides mostly equivalent service to streetcars. Brooklyn-Queens is getting the Interborough Express at some point.

https://new.mta.info/project/interborough-express

From discussions in Oslo, Norway: every researcher or institute says that buses are both cheaper and significantly more flexible, and should be used. And then the public comes along, preferring trams by a mile.

Interesting. Is there a good entry point into the bus vs tram cost and flexibility research?

> They never actually tore up most of these lines; the city just paved over them.

Any place there is a comprehensive utility construction project, there is a pay item that orders the contractor to excavate the rails out of the way for utility installation. A survey is done using a meta Detector to find if any rails remain in a site.

> They never actually tore up most of these lines; the city just paved over them.

I spotted one of these in July in long island city by vernon blvd while they were repaving. You can see them embedded in the cobblestone. Here's some impromptu phone pictures: https://imgur.com/a/MLdjvxo

We are excited for Open Data users to dig into this dataset, experiment, and find insights from the “speed sample of NYC’s streets” that the MTA’s 4,900 buses collect each day. This data will be uploaded on a monthly basis, and can be found on the NYS Open Data portal.

What a great dataset and effort to allow for further research into areas of the city that could benefit from anti-congestion measures (cough cough car tax) to improve bus services!

Also - where does one even store 4,900 buses in NYC? I guess most of the fleet is out on the streets all day, but I imagine servicing all of those is quite the feat.

https://en.wikipedia.org/wiki/Bus_depots_of_MTA_Regional_Bus...

I was sure it would be off the island somewhere so I looked it up. There are many depots around the boroughs and they seem to handle their servicing internally there.

There is one not far off of Times Square.

Exactly this. Lots of bus depots all around NYC. Several across Staten Island alone.

My uncle worked at the MTA for his entire career as a bus mechanic. My favorite story was when a rep from a company selling "green" buses was visiting. Rep said "These buses never break! You guys might be out of work." and then asked "So when do the buses stop running for the day?"

He was less confident in the reliability of his product when he learned the buses in NYC don't stop.

So this is something I had vaguely always wondered about NYC. Is there actually enough space to park all the personal cars, Or is it assumed that some percentage are always on the road. I didn't hear about serious parking congestion during covid so I now assume that yes, there do seem to be enough car spots.

Parking isn’t too much of a hassle in the outer burrows compared to Manhattan.

Also lots of cars get driven in from outside the city, so presumably there were fewer of those during Covid.

Yeah I think commuters are the biggest reason. It's fairly easy to find parking in Manhattan after 9pm. At noon though? Good luck.

There is plenty of parking space for the cars that you see on the streets of New York. It's just mostly in the outer boroughs, on Long Island, or in New Jersey.

What is "all" the cars? Less than half of NYC households own a car[1]; if 100% of households did (or owned more than one, as is common outside of NYC), the city would have nowhere near enough space for them.

(The city already has virtually no space for the 2 million cars that are owned by the city's residents, plus the millions that enter the city daily. We have laws on the books that are intended to reduce the number of unnecessary car trips in the city, but our feckless state leadership has decided that it doesn't need to follow already-passed laws.)

[1]: https://www.hunterurban.org/wp-content/uploads/2024/06/Car-L...

NYC has one of the best open data portals out there. Kudos to all the teams and agencies who manage it.

Shoutout to Philippe Vibien for creating “NYC Subway Stringlines”, one of my favorite (and certainly most used) data visualizations ever. Made possible by the MTA’s GTFS RT feed.

https://pvibien.com/stringline.htm

Note: If you’re checking this out around 6 PM EST, look at the E train to get an idea of what a bad night on the subway looks like.

Each line on the graph represents a train with the Y axis as stations and the X axis representing time. You can follow the trip of the train and get an idea for how well the line is running based on the straightness of the line. If you see areas where the line is flat in the Y axis, you know that a train is being held at a station.

Here’s an example where “stringlines” provide information that a countdown clock couldn’t convey: https://i.imgur.com/u5VGqH4.jpeg

Because the “line” is not progressing past 5th Ave/53rd st, we know that that is where the issue is occurring. A countdown timer would simply either say static or start adding time, but you wouldn’t know how far the next train is from you.

Here’s another example: https://i.imgur.com/mrvrbUt.jpeg

What I can glean from this is that the E train is running with much lower frequency than it was an hour ago, so I should expect longer wait times.

It’s truly a marvelous invention.

I also love this visualization and remember being blown away when I first saw it!

Two notes: 1. These “stringlines” are also known as Time-Space Diagrams in the transit industry, and they’ve been around for a while. e.g. https://www.researchgate.net/figure/Time-space-diagrams-of-t...

In fact Vibien cites as inspiration the official NYCT stringline paper: https://www.worldtransitresearch.info/research/5936/

2. I’ve noticed that at least on the A, the viz is inaccurate? It’s missing a lot of trains.

Found a nice video digging into traffic time-space diagrams: https://youtu.be/E_tk6BGhYDE

There's a nice a-ha moment when he shows aerial photographs.

I knew he cited a NYCT paper on the page itself, but I've never seen the paper so I didn't know how similar they were. Regardless, making this available to the public is laudable in-of-itself. Thanks for the sources, will check out later!

MTA seems to be seriously upping their technology game over the past few years. Anecdotally, I've noticed their apps seem to have better UX and are more reliable. I'd be curious to learn what organization changes have happened to enable this.

Bay Area transplants, like the gays, are a great addition to any community

I'm curious too.

The embedded video does not work for me on Android (neither chrome or firefox, I think it is because it adds an iframe with a /embed/ link with autoplay via javascript after the "Watch this video" button has been pressed)

This is the youtube link: https://youtube.com/watch?v=MsHGqVuIK5g

In case anyone wants to do a similar visualization on their own, this looks like a screen capture of kepler.gl which is an open source web tool for this kind of map data exploration.

maybe someone will finally highlight how ridiculous the gridlock is on the b44-sbs route, particularly through south williamsburg. I regularly see convoys of 4-5 buses arriving at the same time because the traffic through that neighborhood is so bad that the buses eventually catch up to each other and I regularly have to wait 30+ minutes for it on either end of the route.

Just downloaded all the data as a 2.45GB csv file. It took about 10 minutes to "export", before the download started.

Seems like a Parquet or SQLite file would have been more appropriate

Do you need the data to know that busses are insanely slow?

Never underestimate the bandwidth of a double-articulated bus filled with people going to work/play/shop.

Also, slow = safe. Around here, the operators are cautious, diligent, and the best drivers on the road.

MTA bus drivers are maniacs. I mean, I appreciate that they are but cautious or diligent aren’t the adjectives I’d use.

> Never underestimate the bandwidth of a double-articulated bus filled with people going to work/play/shop.

Let's imagine a chart, "Average door to door journey time experienced by a passenger" and "Total journey time" as a function of "Number of people on the bus."

Do you think these lines go up and to the right, or down and to the right? If it goes down and to the right, do you think the slope is steep enough so that the total door to door journey time goes down?

Who is the marginal bus passenger? Is it people who are in a hurry, or who can take their time to get to a destination? What does that say about average door to door journey time?

I don't get it. It seems like common sense to me that busses are slow. As they get used more, they get slower, not faster. There is so much adverse selection for a bus passenger: the very first passenger is for whom it is most convenient, and the very last passenger likely has the worst journeys.

In your scenario, you asked to estimate. The expected average passenger and aggregate journey times of a double articulated bus filled with people is much slower than an empty bus. That said, anyone is entitled to strongly held beliefs about anything.

The only intervention that lowers average and total door to door journey times makes busses look more like Ubers than trains.

It's hard to overlook bandwidth/latency tradeoffs when you are yourself the packet.

Especially when you're a low-latency packet with minimum payload who benefits from a lack of congestion on the network...

Consider a city without any buses: everyone who's drunk, poor, has no license/insurance, underage, distracted by infants or smartphone, they're all on the road with you, walking, biking, in individual, poorly-maintained cars. You may begin screaming for them to get on a bus already.

FWIW I regularly take the bus and find it to be comparable or better than the train for many of the routes I regularly travel. I do have to be more mindful of traffic, but rarely find myself thinking the bus is slow

Am curious as to where this happens. Everywhere I go around the world, train is much faster than buses (a general rule is 2x faster but express trains can be even faster). Anytime there are buses replacing train services, the bus is often at least 2x slower than the equivalent train service (and sometimes they even end up skipping certain low-usage stops for the bus to try and achieve a manageable timetable for the buses). I've not seen anywhere in the world where buses are faster.

As long as you don't make the mistake of trying to take the bus through a neighborhood which contains a community that celebrates a raucous holiday on the day you are traveling. Especially embarrassing when it's a holiday you're familiar with, and you just didn't think about it. But I'll admit, that's a niche complaint.

It appears the purpose of this post was primarily to encourage others to explore the dataset, not necessarily to state, "Buses are slow".

It's possible some "bored data nerds" may find some patterns that help real people in real life while poking around on a Thursday afternoon!

Yes, you do

From the context (buses mentioned) I'm assuming this about the Metropolitan Transportation Authority? I had to search for how the "MTA" abbreviation can be expanded. My mind resolves "MTA" to "Mail Transport Agent".

Reply on news.ycombinator.com