Free Your Data
Evidence is mounting that unlocking your raw data is a great way to enhance public services at very little cost.
One thing government has in abundance is data. Most of it is buried in agency computers. What would it do if it were set free?
Evidence is mounting that unlocking your data is a great way to enhance public services at very little cost.
The trend really started about two years ago in Washington, D.C.
"Our philosophy is simple really," says David Strigel, the program manager for the D.C. data warehouse. "We take data from the agencies, put it in a central repository and share it in as many ways as possible." The well-known success of the Apps for Democracy program was enabled because the data was available in a form that developers could use.
In D.C., making raw data publicly available has lessened the burden on city infrastructure. Since its implementation, city administrators report less time spent fielding questions and requests for information. Data feeds also serve as the informational backbone for the city's CapStat program, an internal performance management system used by city officials to track agency performance against established goals.
The best way of providing citizens with the information they need appears to be to "free the data," and let developers combine and present information in ways government never imagined.
Particularly in tight budget times, public agencies lack the technical resources to present information on all the myriad platforms that citizens desire.
Not long ago, we wrote about the success of the Massachusetts Bay Transportation Authority (MBTA) in releasing data that developers then used to provide real-time bus location information for commuters. One key to that success was deciding not to deliver the information in a traditional form.
For years, GPS data about the location of MBTA buses and trains was available internally, but not available to commuters. It was a change in philosophy, not any technical breakthrough, that freed this data and allowed it to become useful information for commuters.
In the past, the holy grail for real-time transit information was those light emitting diode (LED) display screens located in stations that say, "Next Greenbush Train in 7 minutes." The "cool" transit systems like BART and Metro had them, and everyone else wanted them.
But getting them up was harder than it looked. Josh Robin, Director of Special Projects for the MBTA, says most agencies approached the problem in exactly the wrong order. "They want to start with the LED signs installed in stations. Then they want to build web and phone applications that enable commuters to access the information from the agency. Then they'll look at granting access to the data."
For government, those first two steps are really hard. The challenges of actually building the hardware and software that will translate the data into usable form isn't something generally done at a transit agency. The MBTA isn't in the mobile app business; they are in the train and bus business. With slow procurement cycles, by the time an application was developed for a particular platform, it would probably be obsolete.
In 2006, a federal study on real-time bus location information highlighted the challenges. The irony is that most of the "next train in 7 minutes" LED signs on station platforms don't do anything to reduce trip time or alter commuter behavior in a beneficial way. Once you are in the station, you are pretty much captive. It's nice to know that your wait will be 7 minutes or 17 minutes -- such information can lower "perceived wait time" -- but you can't do much else except wait.
The real goal was putting information into the hands of commuters before they left to catch their bus or train.
Instead of remaining bogged down in development details, the MBTA decided to simply focus on the data. They didn't realize that it would be as easy or as fruitful as it was. Within hours of making GPS data available for five pilot bus lines, developers were creating apps for a dizzying array of platforms -- at virtually no cost to the MBTA. (The MBTA did pay about $100,000 to NextBus, a private vendor, for some predictive algorithms that allow for robust predictions.)
Both in D.C. and at the MBTA, officials agree that it is better to make imperfect data available immediately rather than wait for perfect data. In D.C., media and other observant citizens identified mistakes in the data, which the District was able to clean up.
Meanwhile, the MBTA still only has about three of those LED displays up. But commuters have something much more powerful: Real-time information, literally in the palm of their hand, about when their next bus or train will leave the station.