Home Public Transit Policy We are talking about gondolas, for some reason

We are talking about gondolas, for some reason

by Benjamin Kabak
Gondolas, now a thing. (Photo via East River Skyway)

Gondolas, now a thing. (Photo via East River Skyway)

Every now and then, an idea comes up regarding transit solutions in New York City that seems so out there in its creativity and so out of the box vis-a-vis the way transit operates here that you have to take a step back and appreciate it. Everyone got the ferry bug a few months ago; then we heard about waterfront light rail; and who could forget when John Catsimatidis threw out the idea for a monorail during his run for mayor? Today, we have gondolas.

Gondolas aren’t a particularly new idea for New York City. The Roosevelt Island Tramway delivers over 2.6 million passengers to one side or the other, and until it couldn’t keep up with maintenance obligations and passengers were stranded in the air for hours, the Bronx Zoo had the Skyfari. Now, thanks to Dan Levy, president of CityRealty, we have the East River Skyway, the latest and greatest in niche transportation for waterfront communities on either side of the East River.

The idea, Levy says, came to him while on a ski trip, and his plan involves three phases that will, he claims, cost around $75 to $125 million each. The gondola system, when completed, would span from south of the Brooklyn Bridge through Dumbo and the Navy Yards and north through Williamsburg, Greenpoint and Long Island City with connections to the southern tip of Roosevelt Island, the United Nations and the South Street Seaport.

In a certain sense, this plan gets to problems with the current transit set-up including overcrowded L trains, a need to serve the southern part of Roosevelt Island, especially with the Cornell development on tap and more capacity across the East River. On the other hand, the alignment is terrible in that it tracks subway lines such as the J/M/Z that are under capacity and mirrors preexisting ferry service. The materials tout a 3.5 minute ride from Williamsburg to Delancey St. at the base of the Williamsburg Bridge, but that’s already something the M train can deliver and with better connections to points north. Even though, as CityRealty site 6sqft noted, “gondola stations can also be sited several avenues in,” that’s only the case because the Roosevelt Tram tracks the 59th St. Bridge. (The site also calls the Lexington Ave. IRT line “less taxed” than the L. Make of that what you will.)

Levy apparently drew his inspiration from international cities as well, something that should be applauded, but his examples leave much to be desired. He cites London’s Emirates Air Line as a comparison, but that’s a 10-minute ride geared toward tourists. Germany’s Koblenz Rheinseilbahn is a temporary structure that serves to move people to a cliff above the Rhein, and Chile’s Telerifico Bicentario remains in the planning stages. Levy’s would be among the most complex in the world and relatively long as well.

As to ridership, he predicts around 5000 per hour — which is the equivalent of about three peak-hour subway trains. It’s a reasonably decent ridership, but it’s also one limited by geographic constraints. As with ferries, these gondolas get people from one coast to another, but not from where they live to where they work. Sure, some people live on the Williamsburg waterfront and work near Wall Street, but many would still need to ride a crowded subway. Thus, the problem for which Levy is trying to solve remains. Furthermore, these are issues that could be solved with dedicated bus lanes across the city’s bridges or better bike infrastructure. That’s the realistic conversation we should be having.

Ultimately, this is a fanciful idea, but one that’s more pie-in-the-sky than anything else. It can move the conversation though about ways to solve transit capacity issues, and if someone wants to build it with private funding, no one other than NIMBYs with waterfront views will raise much of a stink. (The insurance costs for operating these types of systems though make them cost prohibitive and nearly impossible to run at a profit.) For now, it’s the shiny new toy.

You may also like


jfruh September 17, 2014 - 12:02 am

My favorite is that the site contrasts “CROWDED SUBWAY TRAIN” with “QUICK AND UNCROWDED COMMUTE WITH BREATHTAKING VIEWS”. It’s going to be an efficient new transit system that only YOU will ride, not all those incovenient smelly other people!

Marc Shepherd September 17, 2014 - 9:57 am

I’m neither pro- nor anti-gondola, but it seems like Ben hasn’t even bothered to read the proposal, before trashing it.

For instance, Ben says that the J/M/Z or ferries would be suitable substitutes. A quick glance at the proposal would have shown that the proposed gondola route does not duplicate any ferry or current subway route.

So the gondola is not a proposed substitute for trains and ferries we already have (or even could have). It’s a substitute for building entirely new train lines, which as we know, would cost billions. Ben also suggested bus service as an alternative, but since he got the facts completely wrong about the J/M/Z and ferries, I am not sure of his judgment on that conclusion either.

Of course, it’s laughable that Ben suggests ferries, of all things, given that Ben has been consistently anti-ferry, as he should be. Ferries are generally poor. They get you only from water’s edge to water’s edge, and the connections there are too far away to be very useful.

Mind you, the gondola almost certainly won’t happen, simply because most transit proposals in NYC never happen. But if anything will be done, something that costs hundreds of millions has a better shot than a new train line that costs billions.

Benjamin Kabak September 17, 2014 - 10:02 am

A quick glance at phase 1 clearly shows that the Williamsburg-Delancey gondola would in fact parallel the BMT Williamsburg Bridge crossing. What exactly were you looking at to suggest I got the facts “completely wrong”?

Additionally, where did I “suggest” ferry routes? I merely pointed out, accurately at that, that routing a gondola from Dumbo to Williamsburg to Greenpoint to Long Island City to the East Side pretty much mirrors the East River ferries. If you’re having trouble seeing how that too is accurate, I can’t help you.

And what a nasty comment in general.

Bolwerk September 17, 2014 - 10:57 am

I hate to take his side on this, and the comment was nasty, but I have to agree this isn’t exactly a duplication. There is no station at the waterfront in Williamsburg, and the walk from the waterfront to the nearest J/M/Z station is probably at least half a mile – (for many people) in the wrong direction!

Yes, it’s parallel, but it doesn’t duplicate the service anymore than the Q60 on Queens Blvd duplicates the 7 Train.

Benjamin Kabak September 17, 2014 - 11:03 am

It depends where you can situate the Brooklyn terminal for the gondola. The rendering places it in the middle of Two Trees’ Domino development which isn’t going to happen. I assumed it would then have to be in the only available space — which is essentially at the end of the Williamsburg Bridge. I’m skeptical even that’s possible for the infrastructure required to support a gondola terminal. At that point, it’s a block or two away from Marcy Av. but also in a very popular bus terminal. Otherwise, I don’t really see the land available to launch gondolas toward Manhattan and receive them from the Navy Yard. That’s the long-winded reason why I considered it duplicative of the BMT.

Bolwerk September 17, 2014 - 11:18 am

Well, there seem to be a million routing issues. But I don’t know if landing at, say, the end of Broadway is conceptually impossible.

I’d actually think a little more of the idea if it just avoided Williamsburg and served the Navy Yard and maybe Manhattan waterfront. This at least means serving a job center, education center, and residential area that all have really terrible transit access.

LLQBTT September 18, 2014 - 2:02 pm

There’s a DOT facility under the Willy B that could be moved or other ”eminent domain” options near there.

j.b. diGriz September 19, 2014 - 1:52 am

You seem to have confused this forum with a Gawker media site.

Ralfff September 17, 2014 - 12:56 am

With so many good ideas, New York will be a World Class City any day now?

Larry Littlefield September 17, 2014 - 7:32 am

The good thing about gondolas and ferries, compared with limousines with chauffeurs, is that no one can suggest that limousines with chauffeurs should be subsidized by the public with a straight face.

Bolwerk September 17, 2014 - 9:38 am

Why suggest it? Just do it.

Jonathan R September 17, 2014 - 7:53 am

I must chime in as I am guilty of advocating for gondolas here in SAS comments. I would prefer to start with connections between the Bronx and Manhattan, which are shorter and could be used to replace bus lines, instead of paralleling subway lines. Consider a gondola over West 145th St that would go from St Nicholas Ave east, across the river to Grand Concourse.

Travis M. Nelson September 17, 2014 - 9:11 am

I spent a week in Zaragosa, Spain in March on business. They have a gondola system there that was built for Expo 2008, to take visitors from the main train station to the Expo, but which supposedly could also be used to bring people from one side of the river to the other after the Expo was over. Now it sits wholly unused. Ironically, one of the themes of the Expo was sustainability.

pete September 17, 2014 - 9:23 am

The gondola will be as fast as a cross town bus. Nobody will use it except as a tourist attraction.

Peter L September 17, 2014 - 8:28 pm

Careful – the ski lift’s maximum speed is 17 mph (or whatever); that’s not an average. Add in the time needed to stop and start and the dwell time, and it will be a lot lower … on average.

As I said, this is just a distraction like those other boondoggles (monorails, etc).

Justin Samuels September 18, 2014 - 12:13 pm

People use crosstown buses all the time North of 59th Street. And they are crowded.

Mike September 17, 2014 - 9:28 am

Medellin, Colombia has an extensive gondola system that is fare-integrated with the elevated train there. It works well.

The proper comparison isn’t to subways, which have higher capital costs and higher capacity, but to ferries, which have low capacity and much higher operating costs than gondolas.

Alex B. September 17, 2014 - 1:59 pm

The key thing about Medellin is that the gondolas there operate in a geography where no other transit is practical. Those gondolas get used for the same reason that the technology works well for ski lifts – it’s the best way to rapidly transport people up the side of a mountain.

The same can be said for the Portland Aerial Tram – it connects a hilltop hospital to transit at the river bank. The alternative bus route is long and circuitous. There are (and will be) strong anchors on both ends of the service. It’s a decent application of the technology.

Dave M September 18, 2014 - 2:56 pm

If you look at the whole proposal it would be a time saver for those communities located near the terminals. Furthermore, the link across the navy yard would save a great deal of time since a bus goes around the yard while the gondola goes straight over it.

If you look at the subsidies given to run buses and applied those same subsidies to a gondola operation you may find the cost not much different.

Nathanael September 20, 2014 - 7:18 am

In keeping with this, gondolas might be the best way to get more traffic across the Hudson River. It’s a big pain to dig tunnels under it, and it’s even harder to build bridges over it, and the ferries are slow.

And, speaking of mountain geography — THE PALISADES. A gondola could run straight from Manhattan to the TOP of the Palisades.

This, of course, will not happen because it involves going between New York and New Jersey, and therefore has no political support from politicians on either side. *Sigh*

Bolwerk September 17, 2014 - 9:54 am

Generally a gimmick, but I can see the gondola making sense for something like the Navy Yard. It even has a secondary benefit of serving the artsy community around Pratt.

j.b. diGriz September 19, 2014 - 1:49 am

I agree, I was thinking the same thing. IF it was built to terminate at a platform built into a J train elevated station.

SEAN September 17, 2014 - 10:21 am

Surprised no one has mentioned the tram that connects the South waterfront area to the main campus of Oregon Health Sciences University in Portland. It’s similar to the RI tram but a bit longer.

SEAN September 17, 2014 - 10:41 am

Portland Aerial Tram

Peter L September 17, 2014 - 4:31 pm

Similar to the Roosevelt Island ride, but lower capacity (78 pax vs 110). Three minute trip with a one minute dwell time means 15 trips/hr * 78 pax/car * 2 cars = 2340 pax/hr … which is about 2/3 of their total daily ridership. IOW, it’s hardly used.

You basically can’t use this type of system on the routes proposed by the guy in the article because you can only have two cars! Double the trip time, halve your capacity.

If you go for the SF Muni-style system where the gripman pulls the handle to stop at a station, then you can at least have a decent system, but I haven’t found an example of a system like that that has cars that carry as much as a bus or two (like Portland and Roosevelt Island).

SEAN September 17, 2014 - 8:30 pm

Keep in mind this tram service I linked to is for medical facilities & medical schools as apposed to pure transit in the case for the Roosevelt Island tram. So the numbers quoted aren’t truly apples to apples. Truthfully I’m not even certain if anything is comparable to the R I tram at all.

SEAN September 17, 2014 - 8:47 pm

[youtube http://www.youtube.com/watch?v.....8;h=315%5D

A ride on the Portland Aerial Tram.

Eric September 18, 2014 - 8:29 am

“IOW, it’s hardly used.”

By which you mean that each car has about 10 riders at a time, rather than the maximum 78.

But 10 riders per car, for a gondola car which does not need a driver, is probably extremely cost-effective. How many bus routes in Portland average 10 riders per vehicle along their entire length? And that’s for a vehicle with a driver.

Spendmor Wastemor September 18, 2014 - 10:26 am

Doesn’t need a driver, but we have the on the RsIs gondola anyway. NYC likes to burn money.

One could have more than 2 cars, if they circled at the ends like a ski lift. Of course, when the thing needs to stop for a couple minutes over the river, buzzards lawyers descend on the pax, and before violating the taxpayers.

John-2 September 17, 2014 - 11:09 am

If you don’t get the gondola termination points in Manhattan close to subway stations, you don’t really have a chance to get year-round customers commuting to/from work, because they’re not going to schlep half a mile or more from the river in bad weather (only the just-reopened Montague Tunnel and the adjacent 1 train have a station anywhere close to the East River, with the South Ferry-Whitehall stop). And if you extend it at least to Second Avenue, as the Roosevelt Island tram is, you are going to have NYMBY protests from people living along the streets where the gondolas would be, who don’t want commuters floating by their windows every 5-10 minutes.

The latter is why putting one next to the Willie B makes sense, because as with the one over 60th Street, you eliminate the problem on at least one side of the street. Only at 96th Street and (if Phase III of the SAS is ever completed) 34th Street could you terminate a gondola close to the river to avoid NIMBY fights, while at the same time getting the areal tramway’s destination point reasonably close to the subway.

Chris C September 17, 2014 - 12:27 pm

Anyone who cites the Emirates Line in London as a good example of this sort of thing needs their head examined – especially if they are proposing it as a possible commuter route.

It went way over budget and despite being sponsored by Emirates it still needs subsidising by the tax and fare payer via TFL – despite Mayor Boris saying that it wouldn’t need a subsidy (much like the Boris Bikes that would be paid for from sponsorship and usage fees but still needs a subsidy)

A 2013 usage report stated that only 4 (that’s FOUR) people made regular use of it such that that got a commuter discount.

And when this fails (which it will) who’ll get stuck with the costs of dismantling the towers etc?

London Banana September 22, 2014 - 6:34 am

When I read “like the Emirates airline” I came on here to comment thus:


the thing is a joke, they made the business case while the Jubilee line was out of action most weekends due to upgrades, on the basis of serving the O2 Arena. But it opened after the Jubilee upgrade was completed. #planningfail

lawhawk September 17, 2014 - 1:02 pm

I don’t think this proposal counts for anything more than a developer trying to gin up more business and interest in his properties situated on/near the waterfront and looking for ways to drive more people to them. In other words, it’s a solution in search of a problem since ferries and existing subways/mass transit are already close to the points where he’s proposed

That said, gondolas might pose a solution not along the Brooklyn waterfront, but for Staten Island to Brooklyn/Manhattan and Brooklyn to Governor’s Island to Manhattan. It might break the logjam on transit projects on Staten Island by creating a faster and more direct routing between central Staten Island and Manhattan, though I think even that is an illusion given that aerial trams/gondolas max speed is probably around 10mph.

Chris C September 17, 2014 - 2:43 pm

Could you imagine the number needed and their size, height and depth of the towers needed to support the cables to enable ships to pass underneath.

Let alone the safety issue if a gondola got stuck or something else happened over the bay?

lop September 17, 2014 - 3:30 pm

You have more than a 100k daily trips between Manhattan and SI on express buses and the ferry today. Build a rail line.

BoerumBum September 17, 2014 - 1:11 pm

Anyone else think this is a fake proposal, planted in the media by press release to help publicize CityRealty, and do nothing else?

Peter L September 17, 2014 - 8:30 pm

Huh. Interesting idea. Wouldn’t surprise me.

Chet September 18, 2014 - 9:42 pm

Now that’s thinking outside the box!

Rob September 17, 2014 - 3:37 pm

So Levy cites London’s Emirates Air Line – a facility that barred anyone with a connection to Israel. Is he proposing to bring more of that kind of antisemitism to NY, or is he just oblivious?

Chris C September 17, 2014 - 5:23 pm

That’s not correct.

There were some clauses in the draft contract based on the fact that Emirates is owned by the Dubai Government who do not recognise Israel but these were removed.

As a publicly operated facility it would be illegal in the UK to try and bar anyone from using it just because they have a connection to a particular country.


Peter L September 17, 2014 - 4:20 pm

Ultimately, this is a fanciful idea, but one that’s more pie-in-the-sky than anything else.

The problem with this East River Ski Lift is just like the problem with monorails, maglev planes, hyperloops, hydrogen cars, self-driving cars, buses that travel over a traffic lane on stilts, blah, blah, blah … They divert the conversation from actual solutions to crap that will never be built (or if it is, will be an anachronism within years, cf. Sydney and Seattle monorails – even Disney runs a fleet of buses because their monorail can’t keep up).

Your bus-lane-on-the-bridges thing is one of those real-world solutions that never gets the kind of traction as these boondoggles, even though they handle 10x the traffic for x/10 the cost (how many passengers come through the Lincoln tunnel in buses again?).

Joey K September 17, 2014 - 9:45 pm

Self-driving cars do not belong in that grouping. Otherwise, yeah, you’re right. This isn’t a serious proposal. If it were, maybe the route wouldn’t cross over the Williamsburg Bridge for no reason. Or perhaps it wouldn’t state that it would create “zero emissions,” as if New York’s power grid relies exclusively on wind and solar energy. How about the claim that it will move 5,000 people an hour in both directions, but then the rendering shows a cabin with seating for six. So what does that work out to, 800+ gondolas per hour?

This is what happens when you have the money to pay a graphics guy to mock up that idea you had when you were in the shower.

Nathanael September 20, 2014 - 7:26 am

Self-driving cars are fantasy material. Here’s why.

The main, permanent, non-replaceable use case for cars is on crappy rural roads in poor weather, and self-driving cars just don’t even have a *chance* of ever being possible in those circumstances; nobody’s going to solve the pattern-matching problem in our lifetimes.

Humans are excellent at matching strange, unexpected patterns and computers are awful at it. So we can tell when there are deer by the side of the road and we know to slow down. We can tell when there’s something wrong with the pavement indicating potential flooding. And we can react appropriately. Et cetera. Nobody will ever program a computer to handle this infinite variety of situations.

You can try to make self-driving cars work in manicured roads in cities, but at that point you have a problem with congestion, and you find that you need buses or trams or subways instead. I suppose they might provide a partial replacement for some taxis, but they’ll always be terribly niche.

Joey K September 20, 2014 - 11:33 am

The technology that exists and is in use in Nevada and California for self-driving cars today is already quite robust. To say that technology will not exponentially eclipse this current version of a self-driving car in just ten years is short sighted. Think of all of the technological advances that you use every day today that did not exist just ten years ago. You will see a mainstream transition over the next decade from traditional operation to computer-aided operation, and eventually, to self-driving options for certain types of trips. I am not a big fan of Google, but what they are doing in this realm is really amazing stuff.

Nathanael September 24, 2014 - 8:00 pm

No, it isn’t robust. I’ve been following the engineering articles.

It’s incredibly fragile. Basically, it does not work.

Nathanael September 24, 2014 - 8:01 pm

And, yes, the engineers have admitted that they are no closer to solving the problems of driving on country roads
(a) at night
(b) in snow
(c) in heavy rain
(d) with bridges out
(e) with collapsed pavement
(f) with detours which aren’t in their GPS systems
than they were before.

Nathanael September 24, 2014 - 8:04 pm

If you want to learn more about why this is never going to happen, look up the “pattern matching” literature.

In short, this sort of pattern matching is something humans are actually *incredibly good at*, and we don’t know how to make computer algorithms which are even a hundredth as good.

There are a lot of tasks where humans are terrible at them and it’s easy to make computers which do much better.

Driving a car in poor rural road conditions, spotting hazards, and reacting appropriately, is a task where humans are actually extremely impressive.

Nathanael September 24, 2014 - 8:08 pm

(For other examples of “computers suck at this”, see Amazon’s “you might like this” recommendations, your bank’s automated “fraud detection” system, and yes, Google search. Astounding amounts of money have been poured into all of these, and they’re all terrible.)

Eric September 21, 2014 - 6:29 am

You can’t get a taxi to drive you out to a “crappy rural road” either. But the vast majority of us use cars primarily for work/shopping/errands on well-maintained urban roads. Taxis work just fine in such an environment except for their cost. And since the large majority of that cost is the driver’s salary, a driverless car would make a much cheaper taxi, which would likely be much cheaper than owning a car.

No, this doesn’t replace the need for transit in dense areas. But it could greatly increase safety, and somewhat increase efficiency, for a large majority of non-transit trips. And in sprawling US suburbs where buses/paratransit are effectively a taxi service already, it would allow for transit to leave and focus its service on core areas.

threestationsquare / Anon256 September 21, 2014 - 12:46 pm

I don’t think driver salary is actually the majority of the cost of taxis at this point. It seems medallion rental alone might be a larger fraction of the price for most NYC taxis (one source gives $7k/month, and since the number of medallions is fixed, one might expect eliminating the driver’s salary to merely drive up the price of medallions). One can hope that the medallion system will be reformed (or loopholed) but it should probably be replaced with less-broken externality taxes like congestion pricing and carbon tax on fuel/electricity. And the cost of the car is still quite significant (especially as self-driving cars will be very expensive to start with).

Self-driving cars will likely decrease the price of taxis somewhat, by perhaps a factor of two or even three, but I think proponents’ expectations of an order-of-magnitude cost decrease are overly optimistic. And 1/3 of current taxi prices is still too high to be attractive for long suburban trips.

Eric September 22, 2014 - 4:24 am

I can’t speak to medallion costs (which shouldn’t exist in a fair market). But the driver is biggest expense for transit buses, and taxis carry an order-of-magnitude fewer passengers per driver, so why shouldn’t the taxi cost be dominated by the driver?

Why would self-driving cars be expensive? They are basically a regular car, plus a computer with fancy software, plus a few cameras and sensors. A computer costs maybe $500, and cameras are pretty cheap. The software is expensive and difficult to develop, but it only needs to be developed once, and then you can bet that it will be priced at a point the market finds attractive.

threestationsquare / Anon256 September 22, 2014 - 5:57 pm

Sufficiently accurate and reliable sensors are expensive. Quoth Wikipedia “Google’s robotic cars have about $150,000 in equipment including a $70,000 lidar (light radar) system.”

I agree that medallions shouldn’t exist but at the end of the day there’s only room for so many cars on the streets of the city, and this limited resource should be priced one way or another. So taxis (self-driving or otherwise) will always have to be significantly more expensive than the subway, because if they aren’t then millions of people will switch from the subway to taxis and traffic will grind to a halt. Whether this is achieved by medallion costs, congestion pricing, fuel taxes or some other means, it will necessarily eat most of the gains from eliminating driver costs.

lop September 22, 2014 - 8:46 pm

The price of equipment is likely to drop significantly over time. That said the cost of an mta bus driver counting pension is near 150k per year yes?

Alon Levy September 22, 2014 - 9:42 pm

Try $63,000 (may or may not include pension, not sure).

Bolwerk September 23, 2014 - 10:52 am

lop may not be that far off. Annual base pay plus healthcare benefits plus typical overtime plus pension costs probably reach into the six figures at some point in the driver’s career, though I don’t know when.

lop September 23, 2014 - 9:45 pm

I don’t see an analysis of how much bus drivers cost the MTA, and would appreciate if anyone had one to share.

2013 MTA bus spent 244 million on payroll, 448 million on labor total. 3576 full time equivalent employees, 2023 are listed under operations: operational hourlies I’m assuming they are bus drivers. There is a separate listing for managers/supervisors. Another 1044 work in maintenance. If the bus drivers only cost the MTA 63k per year and the maintenance 100k per year then the remaining employees would cost the MTA 450k each. That seems unlikely. But alright, so 150k is higher than average for the cost of bus drivers to the MTA, but it’s not going to be that far off, and certainly looks to be over 100k per year per driver.

And a bus has more than one driver – each one isn’t working 12+ hours a day, 7 days a week.

My point was just that even if the cost of driverless vehicles adds 150k to the price of the bus, even if you had to hire some additional employees for revenue control or maintenance on top of that, even if the number of paying riders drops some, it should still pay for itself pretty quick.

Bolwerk September 23, 2014 - 10:38 pm

It’s an absolute bitch to navigate, but payroll information is here:


threestationsquare / Anon256 September 23, 2014 - 12:05 am

Taxi drivers are significantly cheaper, median salary under $40k and no pension.

Nathanael September 24, 2014 - 8:11 pm

Self-driving cars might take over as taxi replacements. Eventually. Maybe.

The default failure mode for a self-driving car has to be to “fail safe” — in short, to stop dead in the road.

The car needs to be able to tell whether the thing obstructing its vision is just a bunch of leaves blowing, or a bunch of papers dropped by a businessman who will run out into the road in a second.

This isn’t going to be a solvable problem. The only way around this is for the car to stop every time the leaves blow.

Nathanael September 24, 2014 - 8:12 pm

…I do think self-driving cars would be a major safety improvement in cities, but nobody is going to tolerate them. They’re just going to turn off the self-driving mode every fall when the leaves start dropping… etc. You see what I mean.

threestationsquare / Anon256 September 21, 2014 - 12:20 pm

If self-driving cars are so imminent why does making a train line driverless require hundreds of millions of dollars and years of disruptive signalling upgrades (rather than just some computer vision hardware on the train), even in places with efficient and competent transit construction/operating cultures like Paris and Japan? Serious question.

Eric September 22, 2014 - 4:28 am

Train wheels have little traction compared to tires, which means that the braking distance is usually greater than the visible range, so no computer vision system no matter how perfect could operate trains.

lop September 22, 2014 - 12:03 pm

They could read trackside signals the way people do.

Bolwerk September 23, 2014 - 11:08 am

Braking distances on subways traveling at speed already seem to be significantly less than the length of a typical platform. That’s typically well within the visual range.

This is probably more of a problem on mainline railroads, especially those that aren’t using MUs.

threestationsquare / Anon256 September 23, 2014 - 4:21 pm

Visual range is irrelevant. A human driver is already a “computer vision system” (and not a particularly perfect one) and the signals are already placed based on feasible stopping distances. Just read the signals.

Nathanael September 24, 2014 - 8:07 pm

Grade separation is the biggest cost involved in driverless trains; nobody’s willing to tolerate grade crossings with automatic train operation.

Much of the rest of it is communicating the signal indications to the train via coded track circuits; older, mechanical systems (such as the NYC subway has, or such as London used to have) didn’t do that. (We do not trust the train’s visual systems to read the signal lights correctly.) If you’ve already got that, the self-driving train is pretty easy to implement.

threestationsquare / Anon256 September 25, 2014 - 1:20 pm

Reading the visual lights (at least as reliably as a human can) really shouldn’t be that hard, much easier than the problems self-driving cars have reportedly already solved (they do at least manage to stop at stoplights under normal conditions). Also don’t most rapid transit trains in e.g. Japan already have in-cab signalling (and grade separation), but still have drivers?

For that matter I’m not sure the grade separation makes that much difference; human train drivers rarely manage to stop in time when a grade crossing is obstructed, so why would a driverless system be any worse? But the busiest lines everywhere in the world are already grade-separated so this doesn’t answer why they aren’t mostly driverless anyway.

Bolwerk September 25, 2014 - 6:23 pm

Automated-but-attended is getting ever more common these days. Though train automation is proven and mature, it seems people aren’t entirely comfortable with it (or unions have enough grip to prevent driverless automation from spreading).

It certainly makes a political difference. It’s common to still need a human being to operate doors, I guess.

(My personal take is having an attendant on the train makes sense. NYC’s practice of having two is rather ridiculous though.)

nicethat September 17, 2014 - 6:22 pm

The supposed routes are incredibly unrealistic. They’d put a gondola tower in the middle of the FDR memorial on Roosevelt Island? And apparently cables over the top of every piece of waterfront real-estate in Brooklyn and Queens not owned by the developer.

RailPhilly September 17, 2014 - 9:25 pm

Gondolas are a bit en vogue right now, but they aren’t necessarily a useless gadgetbahn. They can be a significantly cheaper option to a funicular, for example.

Lining the East River with them doesn’t make sense. Crossing the East River only makes sense if there’s a transfer potential on the Manhattan side, and poor Manhattan access on the Queens/Brooklyn side. Maybe Astoria – Upper East Side, with a stop on northern Roosevelt Island.

Better would be Gondolas for cross-hudson travel, where capacity is desperately needed. Gondolas’ knack for elevation changes could serve both the high ground and low on the Jersey side, clear shipping air draft, and then drop in to manhattan near an Eight or Seventh avenue line stop.

livesinnewyork September 18, 2014 - 12:35 pm

This is an excellent comment and observation. A gondola crossing from NJ serving the new west-midtown development zone is not only a terrific idea but a timely one. Between the huffing about the capacity on the tunnels, needed future closings, etc, a 7-line to Seacacus, etc, this makes perfect sense. The neighborhood is new so no nimby’s – drop a gondola next to the 34th st 7 station and you have a serious winner.

As for the east river i think this is mostly a horseshit proposal to avoid spending on real infrastructure. clearly there aren’t enough crossings to serve the population – we need a new one between Greenpoint and 23-34th st and one from Astoria to UES somewhere.

The Greenpoint connection just screams subway. A new tunnel can connect at least to the Lexington line, have a stop for the future 2nd avenue, connect to the G, and circle through Williamsburg to alleviate the L on its way other under-served areas (Maspeth? Ridgewood? future commercial/office district along route?)

Eric September 18, 2014 - 10:00 am

In the more realistic department, how about a gondola from Hoboken to WTC? No subway or bus lane will realistically get built there in the forseeable future. Or are there airspace/clearance restrictions that would prevent this?

Bolwerk September 18, 2014 - 10:18 am

I don’t know if there are any airspace/clearance issues that don’t apply to the East River, but the PA jurisdiction is probably confounding at best.

(Somewhat ironically, IIRC there was a Calatrava proposal for gondolas from Manhattan to Red Hook and/or Governor’s Island a few years ago.)

Alon Levy September 18, 2014 - 1:26 pm

Hoboken already has PATH. But yes, gondolas would be really useful the Hudson, farther north, either along the GWB or on some alignment that connects to Midtown if spanning the Hudson is not too difficult.

threestationsquare / Anon256 September 18, 2014 - 6:35 pm

Would people really get off a bus at the GWB plaza to transfer to a gondola rather than just staying on the bus to Manhattan? I know the GWB gets congested but it’s rarely bad enough that you’d beat it transferring to a 15-20mph gondola that you probably have to wait in line for. Even for Fort Lee originating traffic the benefit over the existing frequent jitney service (and, eastbound, casual carpools) across the bridge is not that large.

That said I think a more useful crossing further south would be feasible. The Whistler Peak2Peak gondola has a 3km uninterrupted span, and the Hudson is only 1.5km wide.

Nathanael September 20, 2014 - 7:32 am

… I see we’re all thinking the same thought.

I advise Cliffside Park / Edgewater to 125th St or thereabouts for the ideal gondola alignments.

It’s substantially south of the GWB and substantially north of the Lincoln Tunnel.

threestationsquare / Anon256 September 18, 2014 - 6:19 pm

Weehawken or Lincoln Harbor HBLR to the new 34th and 11th 7 station? Or even to 42nd and 8th if it’s feasible to get that far inland. Would greatly increase the attractiveness of both the existing HBLR line and its proposed extension to Tenafly.

On the other hand, the highest-capacity existing gondola systems can only carry around 3,000 people (i.e. two full subway trains) per hour at most, so the lower cost of adding trans-Hudson capacity this way has to be weighed against lower benefits. (About 100,000 people total cross the Hudson inbound in the 8:00-9:00 peak hour currently.)

Joey K September 18, 2014 - 6:42 pm

Why not cut out the connection to a single route gondola and just build an HBLR tunnel to Manhattan? There has been increasing ridership on the HBLR, substantial residential growth in Hudson and Bergen counties, and serious talks of expansion north to Tenafly and west to Patterson. Seems like a Manhattan connection may be the next step for HBLR. (Yes, I know it would be very, very expensive…comes with the territory)

lop September 18, 2014 - 9:46 pm

Would they fit in the Lincoln Tunnel?

Bolwerk September 18, 2014 - 9:56 pm

They should as long as there is no vertical clearance problem.

Joey K September 18, 2014 - 10:40 pm

The Lincoln Tunnel’s clearance height is 13 feet and the HBLR car height from the top of the rail is a little over 13 feet. However, there is no reason why future rolling stock cannot be modified to fit in the tunnel. I think the bigger issue would be scheduling, timetables, Manhattan routes, etc. It’s probably better for it to utilize a dedicated rail tunnel.

Bolwerk September 19, 2014 - 12:15 am

Does that include the pantograph?

It might be possible to recess the rail into the pavement.

lop September 19, 2014 - 12:57 am

Don’t you have several feet above the ceiling and below the pavement in the tunnel for ventilation? Maybe they could close one of the tubes to non-electric vehicles if they need the extra clearance. 8-9am the Lincoln tunnel brings only 3,711 people into Manhattan in autos, vans, taxis and trucks. Shouldn’t be hard for HBLR to make up for the lost capacity of one tube.

Bolwerk September 19, 2014 - 3:33 pm

I don’t know. That’d be fine with me, but it’s obviously a harder sell politically.

Bus/truck clearance and LRV clearance aren’t necessarily analogous anyway. Might be possible for the LRVs to take a center lane and still have clearance because of the tunnel curvature (or something).

lop September 19, 2014 - 5:20 pm

The ceiling is flat. The tunnel itself is curved, but that’s hidden by the ceiling and ventilation system above it (and below the pavement). There is no middle lane, each tube is only two lanes.

Something like this.


I know the Eisenhower tunnel (west of Denver) used to have a big problem with trucks setting off the alarm on the height sensor, often from snow on the top of the truck, not the vehicle itself. They were able to add the few inches of clearance the trucks needed by changing out signs in the tunnel for lower profile ones. If that hasn’t been done in the Lincoln tunnel maybe it would get them enough room.

Bolwerk September 19, 2014 - 7:23 pm

Oops, center tunnel. But I thought the Lincoln was curved throughout. The Holland is more boxy/flat, I think.

Either way, the point is the same: the LRVs should be able to tolerate having only several inches of clearance.

lop September 19, 2014 - 8:44 pm


It’s only curved right at the beginning.

threestationsquare / Anon256 September 20, 2014 - 5:23 pm

If we’re talking about taking over Lincoln Tunnel tubes it makes more sense to devote them to buses, which already carry 340,000 people through the tunnel every day (compare HBLR’s daily ridership of 44,000) and which the tunnel is already pretty much ideally suited for without any costly construction.

lop September 19, 2014 - 12:59 am

If by dedicated rail tunnel you mean new construction then wouldn’t it be a huge waste compared with sending the 7 or L to Hudson county? HBLR to Manhattan might only make sense if you can use one of the lincoln tunnel tubes.

Joey K September 19, 2014 - 9:23 am

If it were just a HBLR tunnel to Manhattan, agreed, huge waste of money. But we are looking to build new cross-Hudson tunnels anyway. We should at least explore the possibility of incorporating multiple modes of transit through single tunnels. Look at the 63rd Street tunnel. Even though there is no sharing of track, there is still multiple uses for a single tunnel construction. The FRA has allowed shared track for light and heavy rail before, this would just be a matter of figuring out logistics, timing, signaling, power, etc. But it is certainly feasible for HBLR to be incorporated into any cross-Hudson tunneling in the future.

marv September 19, 2014 - 1:04 pm

the issue is simple:

What is the net cost cost per passenger trip into Manhattan for each of the proposals to add capacity between NY and NJ.

Net cost equals:
the cost of construction, operating costs and interest costs less the additional revenue to be paid by users.

This would then be divided by the the number of passenger trips over maybe the next 30-50 years.

The question is what should be built? – heavy rail, a subway/path extension, light rail, bus lanes or some combination there of.

The cost needs to be weighed against the other factors including economic growth (including increased real estate values and r/e taxes) etc. Non economic issues need to be factored including decreased pollution (and related health and global warming issues),ease of travel etc.

lop September 19, 2014 - 5:28 pm

Maybe. HBLR might not take up an entire slot from NJTransit/Amtrak or subway, but it carries far fewer people. One NJTransit train might carry 1500 people. HBLR is two cars at 300 each or is it one car? Could you fit in five HBLR trainsets in the space of 2 NJTransit trains?

And if you are building a four track two level tunnel, wouldn’t the 7/PATH + NJtransit/amtrak give you more capacity then one of those + HBLR?

Bolwerk September 19, 2014 - 10:05 pm

Tunnel throughput probably isn’t a problem, but the low-floor HBLR platforms probably kill that idea. If HBLR is going to go to Manhattan, why not do it the cheap way and put it on the surface? It’s not going to make a huge capacity difference.

If we want rapid transit, may as well just expand PATH or the Subway (or NJT even).

Nathanael September 20, 2014 - 7:36 am

Can they negotiate the curves? & the grades? With the clearance?

If they would fit, it would make perfect sense to convert one of the three Lincoln Tunnel tubes to high-capacity HBLR rail service. Of course, that would involve taking lanes away from cars, so it won’t happen.

The Lincoln Tunnel is already full of buses and still won’t allocate lanes permanently to them. 🙁 So.

threestationsquare / Anon256 September 19, 2014 - 7:39 pm

I think a tunnel for HBLR would be about seven times the capacity of a gondola for about 20 times the cost. A commuter rail tunnel would be about 15 times the capacity of a gondola for about 20 times the cost.

Rough estimates: Gondola $500M (5x the cost of London’s, 10x Whistler’s), capacity 2500 per direction per hour (same as London’s). HBLR tunnel $10B, capacity 18000 per direction per hour (thirty 600-person trains). Commuter rail tunnel $10B (Gateway is projected to cost $15B but has some extras that should be cut), capacity 36000 per direction per hour (thirty 1200-person trains).

threestationsquare / Anon256 September 19, 2014 - 7:44 pm

(The extra cost per capacity of a commuter rail tunnel seems pretty clearly worth it to avoid the need to transfer, aside from what we have seen to be the extreme difficulty of getting that much funding lined up at one time. The extra cost per capacity of an HBLR tunnel seems much more dubious, and I think we have a long list of projects offering better value for money (SAS, Utica, Triboro, some sort of Gateway-lite, etc).)

lop September 19, 2014 - 8:39 pm

The expensive part of these projects is the stations right? With LR you’d be able to run on the street. So I would think that HBLR would be cheaper than a commuter rail project that needs a large terminal underground. Some of the cost of the surface treatment would let people travel crosstown in Manhattan, without coming from NJ. As to transfers, it depends who you are trying to serve. Hudson county residents wouldn’t have to transfer from HBLR to get to Midtown.

threestationsquare / Anon256 September 20, 2014 - 5:17 pm

Running in the street severely constrains capacity.

lop September 20, 2014 - 5:21 pm

Isn’t it stuck on the street on the jersey side?

Alon Levy September 20, 2014 - 11:52 pm

Only for a very short segment. Most of the HBLR runs in dedicated rail ROWs.

Bolwerk September 20, 2014 - 11:06 am

Pay little mind to transfers. Every commuter has his/her own pattern. If you save commuter rail people a transfer, you’re probably inflicting one on bus riders or HBLR riders.

HBLR would need to show it can attract the ridership, and much of that ridership would need to be transfers.

Bolwerk September 20, 2014 - 11:03 am

I don’t know where you’re getting $10B from, unless you’re throwing in platinum tiling along the tunnel, but conceptually HBLR and commuter rail have similar passenger throughput capacity limits. Practically speaking, HBLR might be significantly higher with regard to tunnel throughput because trains can move through on top of each other while commuter trains are moving more at speed. The buses already peak at over 30k people/hour, and HBLR can probably do at least modestly better.

HBLR would be (conceptually) more expensive to operate because every additional articulated tram needs another operator, while staffing requirements are lower on subways or commuter rail.

Alon Levy September 20, 2014 - 11:55 pm

The HBLR trains are shorter.

Also, I don’t know where you’re getting the idea that they can move on top of each other. The practical capacity of light rail systems is about a train every 2 minutes. That’s also the practical capacity of commuter trains with good signaling – maybe a bit more when the tunnel tracks split into multiple platform tracks.

Bolwerk September 21, 2014 - 12:02 pm

I mean they can move on top of each other through the tunnel like, and probably with, the buses. This means a nasty speed penalty, but it perhaps maximizes tunnel throughput since headways become tiny. After passing through the tunnel, the vehicles would diffuse in different directions anyway.

threestationsquare / Anon256 September 21, 2014 - 1:11 pm

What are you basing this on? The inner (loop) tracks at Boston’s Park Street Green Line station turned over 200 streetcars an hour in the evening peak in 1911. (Still small compared to the 922 buses that the Lincoln Tunnel handles in the inbound peak hour today.) I think Bolwerk is proposing that kind of operation (somewhat dubious by modern safety standards, but perhaps comparable safety to the buses). Though with such a severe speed penalty, transfer to the gondola starts to sound attractive again…

Bolwerk September 21, 2014 - 2:25 pm

Yeah, but Boston isn’t entirely analogous since vehicles would never make station stops in the tunnel. I assume what they did back then was berth multiple streetcars at once. He’s right about typical LRT implementations where somewhat longer trains probably move faster than streetcars, but that was not what I was talking about.

I don’t see why low headway HBLR through the tunnel would be less safe than the buses or any slower. There could be serious safety issues at subway speeds and/or commuter rail scale, but even that’s probably possible with computing. It’ll still outrun a gondola, at least if there is a transfer penalty to the gondola.

threestationsquare / Anon256 September 21, 2014 - 5:13 pm

Steel-wheeled vehicles have longer braking distances than rubber-tired vehicles at the same speed.

Bolwerk September 22, 2014 - 10:11 am

I guess it could be a problem, but we’re probably talking about speeds under 20mph anyway.

threestationsquare / Anon256 September 22, 2014 - 5:42 pm

As fast as 20mph? Some of the subway lines never get above 20mph. Minimising distances between trains will mean speeds of more like 5mph. (Does the XBL even manage 20mph at peak currently?)

Bolwerk September 23, 2014 - 11:00 am

I’m not sure. Current service patterns through the tunnel have buses coming through on top of each other very slowly. Not saying that’s best or recommended practice, but replacing some of those buses with trams probably doesn’t impact safety much and boosts capacity at least a little.

threestationsquare / Anon256 September 21, 2014 - 1:36 pm

If we’re talk about far-future fantasies for HBLR, I think it would make more sense to build a new regional rail tunnel (either something like Gateway/ARC, something like Secaucus-Bergen Arches-Newport-Fulton-Atlantic across lower Manhattan, or both), divert all the NJT trains currently using Hoboken Terminal to Manhattan, close and redevelop Hoboken NJT/PATH station, rebuild the Hoboken-approach PATH tunnel to curve west to a portal instead of east, and convert the HBLR lines into branches of PATH using the capacity freed up on the trains that currently go to Hoboken. Something like http://goo.gl/maps/kJdR8 . But the gondola could happen a lot sooner.

marv September 18, 2014 - 11:45 pm

just how many “large” ships actually travel the hudson and east rivers each day?

maybe it is time to relocate the passenger ship terminal south of the battery tunnel in Brooklyn or NJ and then build a series of low/medium level spans with each having a draw span built in across both rivers.

to avoid bringing even more traffic into manhattan, spans should be restricted to: commuter rail, subway, light rail, pedestrian/bicycles, and official/emergency use.

Ralfff September 19, 2014 - 3:13 pm

This is actually a good idea in general, assuming it’s cheaper than a tunnel to build. But as Alon Levy has shown, tunnels in general aren’t actually that expensive, stations are.

MR September 24, 2014 - 11:22 pm

Living uptown I have always thought a gondala service across the GWB could have been a good idea. As the bridge itself was designed to hold more weight than it does and you have the GWB Bus Terminal with a subway stop on the Manhattan side already. On the Ft. Lee side, I do think the ideal situation would be to build a large parking garage with a station to alleviate traffic on the bridge.

marv September 27, 2014 - 10:02 pm

A gondola over the bridge – another time consuming transfer on a high traffic corridor with low volume gondolas!!!

Extend the A train (or the 2nd Ave subway via 125 to St Nicholas Ave)over the (GW) bridge.

Move the bus terminal to cheaper land on the NJ side.

Between bus traffic and commuters now getting on the subway in NJ traffic saved would more than make up for the two lanes lost.

threestationsquare / Anon256 September 28, 2014 - 4:40 am

The bus terminal is on top of the freeway, not exactly prime real estate. Given the absurdly inflated construction costs of infrastructure projects in and around New York, moving it across the river would cost far more than the value of the land. And for what? Everyone would still have to transfer. At most times of day the trains would likely cross the bridge more slowly than the buses do (certainly this is the case on the Manhattan and Williamsburg bridges currently).

If GWB traffic is holding up the buses, stripe some bus/HOV lanes. We can talk about a subway extension if and when those fill up.

(Now, if you could actually extend the subway to Paterson or similar so that many passengers wouldn’t need a bus at all, and upzone the Rt 4 corridor for TOD, that could be a lot more worthwhile. But there are at least a dozen higher-priority subway extensions in and around the city so that one will have to wait until after we can solve the construction-cost-corruption problem.)

threestationsquare / Anon256 September 28, 2014 - 4:47 am

(The same, of course, applies to a gondola at the GWB. If a Hudson gondola is to be useful at all it should be well south of there, linking the very dense neighbourhoods in northern Hudson County, or HBLR, to the UWS or Midtown, rather than duplicating an existing 14-lane bridge. Though it may have a hard time competing with existing bus service even then.)


Leave a Comment