That's partially because cities and counties don't have anything but the barest sense of what routes are in high demand and how outside events affect traffic and they're usually collecting data manually (from those guys you may see on the subway with a clipboard and a clicker, or handing out rider surveys). When it comes to improving transit, they're trying to hit a pinata while blindfolded.
Enter Urban Engines, an analytics startup founded by a two ex-Googlers, a Stanford graduate and a PhD, that plans to use the data users already generate every day to deliver a better, deeper view into cities and how people move around in them. Urban Engines calls its purview the "Internet of Moving Things," and it's all about taking an algorithmic approach to public transit.
The company's roots trace back to when CEO and co-founder Shiva Shivakumar was still at Google from 2001-2010, working on usability and figuring out how people actually move through websites. Over time, observing online user behaviors gave way to wanting to know more about how people navigate the physical world, too, something made easier in an age where location-based data is more readily available in larger quantities than ever before.
Shivakumar made a trip to Stanford University to recruit a founding team: Balaji Prabhakar, a professor of Electrical Engineering and Computer Science; ex-Google systems engineer Giao Nguyen; and Deepak Merugu, who has Bachelor's degree in Electrical Engineering from IIT Bombay and Master's and Ph.D. degrees in Electrical Engineering at Stanford. Urban Engines officially opening for business in May 2014.
The company's mission: to help "navigate the real world as efficiently as the digital one." It's venture backed and funded by firms and investors like Google Ventures, Andreessen-Horowitz, Ram Shiram and Google Chairman Eric Schmidt.
So, back to that data collection. Most of what is gathered by transit agencies is slow, making it hard to get anything resembling a current picture of how people are using local transit systems. Moreover, since each agency tends to keep its findings to itself, it's hard to see how, say, a traffic jam downtown might affect subway usage uptown.
"Train guys know what trains do, bus guys know what buses do," said Prabhakar.
Urban Engines is a little different: By constructing a map based on data provided by transit agencies, it can come up with a working model of the city that gives real-time visibility into what's going on. What's more, it doesn't require any special effort on the part of local government. The company relies on, as a major resource, the data generated by tap-in/tap-out NFC payment cards like the ones used on the Washington DC metro, where Urban Engines is presently engaged (along with Shanghai and San Paolo).
The result is an overhead view of where people are getting on and off of a transit system, from which you can infer a lot of things. For instance, if someone swipes into the system at 5:30 p.m. and out at 6:30 p.m., you know their trip took an hour. But if that's only supposed to be a 40-minute trip, you can safely assume they missed a train or two due to overcrowding. A lot of people swiping in at once A train is coming. A lot of people swiping out at once A train just arrived.
Add buses and traffic data to the mix and you can see how events like school letting out can affect flows around the city and making it easier to identify bottlenecks. For the first time, city planners can get a full view of what's going on in close to real-time, giving them insight into what's needed to make improvements.
"We can figure out what's going on," said Shivakumar.
That information is not just of use to transit agencies, obviously: Urban Engines is in talks with delivery and logistics companies to help them optimize routes and better understand the way things move.
This week, Urban Engines released a mobile app for iOS and Android that brings what promises to be a better mapping experience, due in no small part to the fact that it pre-loads maps and its routing algorithms to the device for 10 U.S. cities, meaning you don't need connectivity or a data plan to access it once the app is downloaded. It also includes a nifty "X-Ray Vision" augmented reality mode that lets you see where you're going through the camera with a map overlaid to minimize the odds you get disoriented.
The idea is that all the data and algorithms Urban Engines is refining are useless if the information doesn't get into the hands of the people who are doing the moving -- and, perhaps more importantly, it gives Urban Engines another sensor to gather more data about those moving things.