Infographic Of The Day: Could Twitter Help Us Create Smarter Transit Routes?

Eric Fischer used geotagged tweets to create maps of the most highly trafficked thoroughfares in major cities.

Traditional city maps visualize just one aspect of urban design—the city’s intended structure, full stop. But add in a layer that visualizes how people actually use the city, and then the map becomes much more interesting. Eric Fischer did exactly that when he used Twitter’s API to collect tens of thousands of geotagged tweets and map them onto the streets of New York, Chicago, and the San Francisco Bay area. The maps amount to something close to a desire path on a macro scale: The maps show where our buses and subways should be, if they conformed to the way we actually move and live.

[In this map of Chicago, Fischer notes that the curve heading southwest doesn’t conform to any transit line—so why should there be one?]

They are essentially inkblot visualizations of where the density of urban flow is highest. (Thicker black lines mean more people passing down that road, street, or mass transit line.) The statistical and mathematical methods used to generate the maps will be of interest to cartographic data nerds (I’m looking at you, Tim De Chant), but laypeople can appreciate the patterns that emerge. In one visualization of the Bay Area, Fischer notes that "Telegraph is the street most neglected by existing rapid transit infrastructure."

[In the Bay Area, Fischer observes that BART, in focusing on downtown commuters, doesn’t do a very good job of serving the corridor between Berkeley and North Oakland.]

Fischer’s maps are yet another example of the kind of powerful information that networked, ubiquitous sensors can provide. The MTA in New York City is spending billions of dollars to expand its mass transit system; how might a map like Fischer’s inform decisions on where, when, and how to dig new subway lines? With regard to the Bay Area, Fischer notes that "BART does not do a very good job of serving the most promising corridor in Berkeley and North Oakland. … Needless to say, if this were to be constructed, it would have to be pretty much entirely in subway to avoid tearing down the neighborhoods it would intend to serve."

[When you look at New York, the central spine of Broadway leaps out at you. But the other big spines are telling as well: They conform to subway lines. In New York, neighborhoods truly live and die by the subway.]

Commenters on Flickr are already beseeching Fischer to offer his maps as large format art prints. In the meantime, you can peruse the details of his maps in our slideshow.

Add New Comment


  • madelinebrozen

    I'd be great if the data behind these maps were open to the public. I'd love to try and play around with other overlays like putting the transit lines overlaid. But also, putting some geographic/demographic overlays. Do you know if this would ever be possible to share in a google earth or some other cartographic format?

  • David

    One thing comes to mind when I see these kind of schemes. NIMBY. If this showed the path going through a rich suburb and/or some politicians back yard, all the tweet maps in the world would push it through.

  • Ben Jones

    I does this take a lack of connectivity into account? I know in Toronto the underground sections of the subway mean that no-one can tweet from that location (no wi-fi or cellular networks on the train). 

  • Guest

    This is interesting, but I wonder if he looked at what the
    ppl were doing while they were tweeting and what time of day it was….

    I mean, if I were on transit, I would probably be tweeting
    because I was bored. You don’t typically tweet while you are walking down the
    street.  Because these transit systems are already in place it would obviously have an impact on where the tweets are coming from.  Additionally, if the tweets tracked were identified as a point of interest only (home, work, restaurant, etc) then you would just have a map that shows the HH, EMP, and Retail densities we already know to be true. Interesting relationship, but coming from a traffic engineer/planner I am not sure this shows the whole picture.

  • JRinks

    This is a really cool data visualization tool and a great use of social media data, but I agree with those other commenters who point out that a) these maps are showing us how people travel now, using the already-existing infrastructure  and b) we can't assume that everybody uses twitter to the same degree.   Could this be re-worked to use trip start and end locations (not the actual route taken) to create trip flow data that could then be used to design transit improvements? 

  • Andrew Chapello

    Excellent visual representation of the actual flow of traffic. If only transit agencies would pay attention to this data when conducting transit planning.

  • Susan Bodiker

    These illustrations demonstrate why transportation routes are called "arteries." They look like angiograms! The point about economic disparities is spot on too.

  • slok

    On a purely aesthetic level, this is a gorgeous visualization.   Bring the organic  to the digital

  • je

    I'm not sure how this maps movement? Aren't geocoded tweets limited to where a person is at any given moment?  It seems like you'd need a high frequency of tweets along a route for a single person to actually get a desire path.  I'd love to hear more about the methodology.

  • Brian G. Dowling

    So do we have any demographic data about the socio-economic classifications of Twitter-users and smart phone owners? As the tweets are geo-tagged and likely as an embedded feature of the phone, the tweet does not have to be related to the location or travel from one location to another it would seem.  Finally, how hard would it be to verify that that actual pattern of physical movement by people matches these patterns?

  • Jesse Kinsman

    I think the issue that this is bias toward a certain social group (i.e. the twitter user) is pretty obvious. If you were to actually use this to plan public transportation routes, you obviously would have to get a better data. 

  • acj

    Agree with Sheena, Fogg, and cg.  Startups often do the same thing by assuming everyone is in line with their own demographic.

  • Sheena Medina

    I agree with cg in that the idea that these maps "show where our buses and subways should be" is not exactly true. If you look at Chicago and New York (two places I have lived for several years), you quickly realize that these maps highlight the economic disparities within cities. The people who are really underrepresented here are the transit users that may not have smartphones. In New York, the farther you get from Manhattan and in Chicago, primarily on the South side are perhaps where public transit is needed and utilized most. If anything, the inverse of these maps show where our buses and subways *should* be-- the places where income and access are low.

  • Matthew

    I noticed this immediately, especially with Chicago's very strong emphasis on the north side of the city, which is generally more wealthy and more white.  This doesn't match population densities at all.  Some combination of population density and economic activity should provide a good mix of residential, work, and leisure destinations.  This just maps where people over a certain economic threshold are likely to be.

  • Jesse Kinsman

    This shows the routes people take to get places but unfortunately it is going to be restricted to already existent arterials as people either have roads or mass transit to get from point a to point b for longer commutes. 

    So while cool, it doesn't actually show the best mass transit routes as the most traveled arterials are going to be the interstate or other highly used arterials, which is not necessarily the most efficient routes for mass transit.

  • Fogg

    Cartographic data nerds tracking techno nerds? I thought public transit was for real people and they don't always Tweet. In the USA, you once had the dark horse presidential candidate win after the fools at a national newspaper polled by telephone—at a time when only the rich had phones—and had even set the wrong Headlines ready to roll as they tried to declare the wrong winner based on on faulty data.
    Nice graphic but isn't that's always how the data crumbles in the hands of designers.
    But what do I know? As one who believes that being connected is kind of like...voluntary slavery.

  • cg

    "The maps show where our buses and subways should be, if they conformed to the way we actually move and live." Hmm, assuming we want our transit system to work better for Twitter users. In NYC, we already put a lot of infrastructure in in these areas and not in the population-heavy but twitter-light neighborhoods like East New York. This is only a good way to plan transit routes if you want to reinforce inequity in cities.