Smart Para-Transit: A New Vision for Urban Transportation

This is the first article in a five-part series by Streetsblog publisher and LimeWire founder Mark Gorton:


is a crushing problem that oppresses our city, yet many people who drive into
New York each day do not have a good alternative.

I’m an engineer by training and the traffic
flow problems facing large cities today have many similarities to the
engineering issues that I have encountered at LimeWire, the peer-to-peer
file-sharing service that I founded. LimeWire
involves many computers connected to each other passing messages around a
network. Early in the development of
LimeWire, the network was choking on its message traffic as each computer tried
to send more messages than the network could collectively handle. The solution to this problem involved having
each computer reduce its message traffic and organizing the network to take advantage
of efficiencies that could be gained by designing a new computer network architecture.

a network management point of view, the road networks of New York and many
other large cities are horribly engineered. The traditional traffic engineering solution to congestion problems is
to try to increase capacity. However, similar
problems in computer engineering are solved by reducing the underlying need for
traffic. Biological systems, which are
the most sophisticated systems on the planet, are extremely judicious in how
they move things around.

surface transportation system today is premised upon the primacy of the private
automobile, yet the private automobile is the single most inefficient means of
moving people in a city.
By catering to
the private automobile, we have inadvertently made an engineering choice that
maximizes danger, noise, pollution, and congestion and creates a host of other
problems that suck the life out of our public spaces.

less than ten years, with minimal capital expenditures, we can create a new form
of mass transit that transforms the way we run our surface transportation
system and drastically reduces the need to have private cars in New
York City.
I call this new form of mass
transit Smart Para-Transit. Smart
Para-Transit takes advantage of innovations in information and communication
technology to create breathtaking increases in efficiency of our road
network. My very rough initial estimate
is that widespread adoption of Smart Para-Transit would allow for an 80 percent
reduction in automobile traffic in New York City.

basket of ideas involved in Smart Para-Transit are too long for one blog
So I am serializing the explanation over the course of a week. For those of you who can’t wait, you can download and read
the full description.

  • Larry Littlefield

    Why not dynamic carpooling? Same idea in that IT is used to create a flexible form of transport, but the government doesn’t have to supply the van and driver.

    The riders would pay the driver the equivalent of a transit fee, and save money by not owning a car (or a second car). The drivers would earn that money in the course of their normal routine. The carpools would be created for each ride among registered carpool members, allowing schedule flexibility (and carpools for non-work trips).

    The carpool organization might have to provide vans as you describe as a back-up.

    My concern is that NYCT paratransit has long wait and trip times at a huge cost per ride.

  • These kinds of plans used to scare me a little. You see I’ve only seen them proposed as replacements for public transportation such as rail and light rail and busses. But looking at the pdf, I see that is would be a whole new mode– a sort of better more useful kind of taxi. It is “para” transit after all… not just “transit.” Working along side a public transport network, biking, and walking this could make a lot of sense. The technology is already in place. Don’t all the cabs have GPS now anyway?

    I wonder what density of vehicles one would need to serve a given region? This is starting to sound like a traveling salesman problem!

  • Ace

    “many people who drive into New York each day do not have a good alternative.”

    Lost me right there

  • JK

    Some of the technical details could be further worked out if taxi GPS is fully exploited, and medallion cabs are allowed to be dispatched via GPS/SMS, and allowed to pick-up more than one passenger. Similarly, for intra-city trips, livery cabs could do the same. Within the five boroughs, the vehicles are already driving around. Given the point to point convenience of private cars with free parking at home and work, it’s hard to see this as competitive without getting rid of placards and employee provided parking and instituting some kind of road pricing.

  • CH

    “Take for example the group of people who want to travel from Tribeca to Montclair, NJ around 5:30 PM on a Tuesday. There might be a dozen people who plan to make this trip by car in a 15 minute period. These dozen people might require 8 separate cars for their trips. Instead of 8 separate cars, one large van could fit 12 people and consolidate these 8 vehicles into just one vehicle.”

    Or… These 12 people could take the train to Montclair. Why is the baseline assumption that all of them would drive?

    It’s a nice academic idea but paratransit (and let’s be clear, this is not a new concept) does require patience on the part of passengers. To assume that you’ll get a quick pickup and straight shot exactly where you want to go is a bit ambitious. The number of vehicles and operators required to make this time and financially advantageous would probably push the system beyond its intended efficiency.

    Paratransit services are considerably more expensive to operate than fixed route, higher capacity services.

  • Adam

    What you have to ask is why are those 12 people driving in from Montclair in the first place (rather than taking the train) and could an efficient paratransit system (which would still be slower than the personal auto) still meet the needs of those people?

    My guess is no. If you’re driving to NJ from Tribeca during rush hours it’s not that there is a lack of options, but that you PREFER the cost/benefits of your car to the cost/benefits of transit. Most other people doing that trip choose transit.

    On the other hand, I would welcome the opportunity to share rides with strangers going my way if it saved me some money, was safe, and had the on-demand (or close to it) convenience of taxis. Being able to do this for non-commute trips would be especially useful.

  • In Lausanne we already have a smart paratransit system (publicar) that connects all suburbs that cannot be reached by bus or train.

    users call a number and reserve their trip, then the bus start its tour, taking all people and bringing them to their destination. It’s not an alternative to train, but rather a complement.

    more info (in french) are on:

  • Marco, do people commute to work on that thing– or is it for holidays?

  • Michael Steiner

    There are also some information in english on publicar (which exists in various other regions in Switzerland as well for quite a while). I have also my doubt that it would work (efficiently and cost-effectively) in really densely populated area as NYC compared to thinly populated areas/feeders to trains as in PubliCar.

  • anonymouse

    It seems to me like there is a good niche for this sort of paratransit: at suburban train stations in places like New Jersey, where it could function as basically a shared taxi taking people between the train and the widely dispersed destinations that are typical of the suburbs.

  • Juan

    I am so excited to hear someone with the credibility of Gorton suggesting this idea. I’ve been thinking about something similar for years, and really see it as inevitable.

    What this proposal doesn’t specify, which I expect to be critical to the implementation of this system is Social Networking.

    Something like Friendster or Facebook is the perfect basis for encouraging individuals to overcome their hesitations about being in a confined space with a stranger. Anonymous people are weird and scary; sharing a morning with a gal who lives in Park Slope, went to Vassar and owns a Retriever seems perfectly reasonable.

    The day you start LimeTransit, let me know. I want to help.


  • gecko

    If I have a vehicle that gets me where I want to go, I just get in it an go.

    If I forget something I can go back and get it.

    If I remember that I have to do something on the way I can do that also.

    If I see something that is interesting on the way I can check it out.

    If that vehicle is not much larger than I am so much the better, especially if I can fit in my small apartment.

    I . . . , I . . . , I . . . , I . . . , I . . . , I . . . , I . . . .

  • Mark, it’s an intriguing proposal and nicely written. Some questions:

    1. Is it intended to replace taxi cabs?
    2. Would the fare be equivalent to, less than, or greater than the same ride by cab?
    3. How would the fare compare to mass transit?

    The example provided (Tribeca to Montclair at 5:30 p.m. on a Tuesday) is for a suburban commuter. As a city resident, the times I would generally spring for a cab are when a mass transit exists but is highly inconvenient and time-consuming. So my typical scenarios are:

    1. Inwood to JFK or LGA. I’ve got luggage, and it’s only 20-40 minutes by cab, but almost 2 hours by mass transit, so the time savings are significant (especially given the additional waits one experiences at airports).

    2. Downtown or Brooklyn to Inwood at 2:00 a.m. on a Saturday. The subway trip involves multiple lines and at best the A is running local, but more likely they’re running a shuttle bus from 168th St north or a separate train. Again, the time savings are significant.

    For my needs, I’m not sure whether Smart Para-Transit or better regional rail service (like the Paris RER) would be the answer. If I could get a Metro-North-type train with a maximum waiting time of 15 minutes from downtown Brooklyn to uptown Manhattan in the middle of the night, that would be great.

  • Urbanis, Thanks for your nice words. I haven’t explicitly thought out the exact relationship between smart para-transit and taxis. I suspect that real world issues will dictate the eventual interactions. However, smart para-transit is conceived as an alternative to most car trips, and taxis are in many ways an inefficient user of street space as private automobiles, so hopefully, smart para-transit could take some demand away from taxis.

    Smart para-transit need not be bad for the taxi industry. In many ways, the taxi industry is well placed to handle many of the issues involved with managing large fleets of vehicles, hiring drivers, etc. But with smart para-transit the vehicle type and vehicle routing would change.

    I would expect that the fare would be much less than with a regular cab. The cost of the trip would be shared by a number of people, so each person would have to pay only a fraction of the total cost.

    The pricing of smart para-transit is a complicated issue. My gut feeling is that the fare paid should be highly correlated with the cost of providing the trip. I suspect that for most trips in NYC, the cost would be more than typical mass transit, but less than a cab or private car (once parking, etc. are factored in).

    The trips you mention to/from Inwood in your comment are excellent examples of voids in the current mass transit network that could be filled by smart para-transit.

    Smart para-transit is not intended as an alternative to Metro-North and other commuter rail. A surface vehicle trip, even a van or a bus, is more harmful to society than a train trip. One potential issue with smart para-transit is that it proves so attractive that it moves some trips away from traditional mass transit. This shift is somewhat inevitable, hopefully, pricing can be set to discourage this sort of mode shift.

  • Urbanis, the Paris RER shuts down between 1AM and 5AM, and in between you have to take the bus. In NYC at least we have train service at that time; if we had express service it would help.

  • Juan, I am delighted to here of your interest. I intend to keep my transportation advocacy and my businesses separate. I prefer to avoid charges of conflict of interest. However, I do see that putting a smart para-transit system in place will require a huge amount of work. Even the advocacy alone will be a huge effort.

    In addition to producing Streetsblog and Streetfilms, The Open Planning Project does a good deal of open source software development on a non-profit basis. I might consider having The Open Planning Project build open source software to power a smart para-transit system. Once this open source smart para-transit software exists, a city that wants to start a smart para-transit program could easily adapt the existing open source software to its particular region at very little expense.

    TOPP already produces some of the components necessary for an eventual smart para-transit software back end. In fact, I came upon the idea for smart para-transit while thinking about traditional transit routing software that TOPP is considering building under contract to a large US city.

    A full fledged smart para-transit software system is going to be a lot of work and cost a lot of money to develop. I am hoping to have TOPP get a development contract from a city interested in deploying smart para-transit to help fund the software production.

    I can also see building facebook applications for ride sharing as an intermediate step to developing a full fledged smart para-transit software suite. It certainly makes sense that a facebook interface would be part of the eventual smart para-transit system. Grouping people with similar interests on trips could make the whole system more appealing.

    The skills TOPP is going to need for this project (if it ever gets started) will include software developers, project managers, and people to help get the software development contracts

  • If a comprehensive Smart Para-Transit system were available, about the only scenario where I’d see individual cabs would remain competitive is for very short trips (say, going from Columbus Circle to 14th St and 8th Ave). Otherwise, it’s likely the significantly lower cost of a shared fare will trump the convenience of the dedicated taxi ride, assuming there are not too many drop-off/pick-up points and wait times are minimal for SPT.

    As for SPT vs. private cars, I would imagine the time inconveniences of waiting and multiple drop-off/pick-up points would be more than compensated for by the time savings of not having to retrieve one’s car from a garage/street and not having to find parking again at the end. In addition, the pleasantness and productiveness for most people of having someone else doing the driving and parking would be a serious benefit.

    For all the touted conveniences of owning a car, I found owning, parking, and driving one stressful, expensive, time-consuming, and dangerous. Any transit system that gets me where I need to go in a safe, convenient, and timely fashion without forcing me to drive (unless it’s a human-powered vehicle like a bicycle) is one that I’d embrace.

  • willis
  • rlb

    About two years ago I had a broken ankle (a fortunately minimal injury after a car ran over my leg), and as I stewed in my hot apartment reading Kunstler’s Long Emergency and probably going slightly mad thinking about what the hell was going to happen to suburbia, I essentially came to the conclusion of Smart Para transit. I personally think of it as Smart Car Pooling, but the idea is all but identical; a person says I need a ride from here to here, an algorithm analyzes the situation and eventually a driver is told where to go using a gps system. Maybe the person is even given an option of pick up times.
    Under the spell of JHK, I imagined driving a car significant distances IMPOSSIBLE due to the expense. Few would be able to afford efficient vehicles. With such conditions people who live in the suburbs would be forced into commuter rail. How do they get there? They drive. But as the revelations spread, the parking lots would become packed (as is happening across the country now) and the price of parking would go up. This is the situation where I thought para transit was ripest – when parking at commuter rail stations becomes prohibitively expensive.
    In this context, Para Transit would be simplified to a small area and a single destination/departure. The trips would be short and inexpensive. A van would make many rounds during the rush hour period and would be off for most of the day. Eventually it could spread to complete trips to the CBD, but those would be costly because they would take a long time and would eat most of the morning for a given van.

  • Streetsman

    This idea seems to work pretty well for those airport shuttle businesses. they can pick you up at your hotel at a scheduled time, along with other people in nearby hotels that have scheduled similar times. works great in cities that don’t have mass transit to the airport. one van with 10 people better than 10 separate taxis.

  • Sound thinking, and all possible. I own a car and would give it and its garage up in a heartbeat if I could get to points out of the city when I needed to or wanted to. Most of those places are not served well by mass transit.

    The information technology described is badly needed in other segments of our transportation infrastructure too:

    • Dynamically-priced parking and standing for loading and unloading (if the “smart” meter knows the supply of space on the street it can price accordingly, and can direct the driver to the nearest spot and charge for it immediately).

    • Bike Sharing (of course!)

    • Smart traffic lights. So that cyclists are not expected to wait for a red light on the Greenway at 2 am, for example, when there is no one in sight!

    • The conventional bus routes could, of course use it, too. No more bunched buses and info on arrival time available at all stops and in the cloud.

    In terms of business models for car sharing, fractional ownership of cars might work even better than the ZipCar concept, that requires the company (in this case Zip) to lease the vehicles, a big capital cost. Fractional ownership is now used by the wealthy for private jet services like NetJets, but could work just as well people of modest means who need access to cars, bikes, pedicabs, even Segways.

    Finally, one should recognize others who have blazed the trail here, including deposed ZipCar founder Robin Chase, whose more recent idea,, is promoting car-sharing in the Boston area today but is available in all cities.

  • Regarding a Facebook application for ride sharing, see – GoLoco was started by the same people behind Zipcar. I’ve got it on my page but haven’t been able to take advantage of it yet. There need to be a lot more users for it to work.


Peer-to-Peer Mass Transit: How to Make it Work

Here is the second installment of Streetsblog publisher and LimeWire founder Mark Gorton’s essay, "Smart Para-Transit: A New Vision for Urban Transportation." Part 1 is here and you can also download the complete pamphlet. Advances in information and communications technology offer the possibility of optimizing the performance of our existing road network in ways that […]

Streetcars and Street Design

This morning on the Streetsblog Network, Yonah Freemark at The Transport Politic talks about streetcar plans in Washington, D.C. Without better street design, he argues, the capital’s streetcars could end up mired in traffic:   (Photo: rocket ship via Flickr) I have documented some of the quotidian situations that will result in delayed traffic as a […]