Open Data, Crowdsourcing, and the Airport Security Fight

Josh Sulkin, a graduate student at the University of Illinois Urbana-Champaign, might be about to get his big break.

Inspired by the Sunlight Foundation and the Apps for Democracy contests, he created FlyOnTime, an app that mashes up data provided by several government agencies to give people an idea of how likely their flight is to be on time. Users put in a day of the week, two airports, and get a statistical breakdown of flights delayed, flights cancelled, and average delays.

Then, in summer of 2009, Sulkin built in an added feature: Using a smartphone app, users can now track their wait times at airport security checkpoints. He's also merging his data stream with that of the Transportation Security Administration's own app to track wait times.

There hasn't been very much interest in crowdsourcing the data, though, and so he is hesitant to provide figures about wait times based on the few hundred contributions he's collected so far using a mobile-enabled website, even when aggregated in with the TSA's data, which Sulkin says has about 11,000 contributions and launched in June 2010. (That data is pushed out through a Twitter feed.)

People who have not been under a rock for the last few days, or at least have MyFi cards with them down there, can surely see where this is going.

As The Transportation Security Administration is defending itself against widespread criticism of its new ultimatum to passengers who are selected for additional screening — submit to either a highly invasive pat-down or a very revealing stroll through a high-tech imaging device — delays at airport security are becoming a political football. Will people opting to take a pat-down, rather than let a new piece of technology generate an image of what they'd look like without any clothes, generate widespread airport delays? Are they already? Who knows. All we've got to go on now are the assurances of various spokesmen, and whomever a reporter happens to snag at the airport:

Some travelers say that they are also seeing the amount of time spent in security lines increase.

"Last week was the worst I've seen in about a year," says Mat Castaneda, director of sales and marketing for an energy technology company in Orange County, Calif. He says that he waited nearly an hour during screening at Los Angeles International for a flight to Houston.

"I am getting fed up, and I think a lot of people are becoming a little more frustrated about the compounding security procedures," he says. "It doesn't seem like they take anything back. It started out with no liquids, and then (you had to) take your shoes off and now you have to basically strip down."

He chose the scanning machine rather than the pat down. "When I looked at the line, I thought it was just quicker," Castaneda says. "I didn't really want to take another five to 10 minutes. ... It was the lesser of two evils."

The TSA says that wait times have not risen because of the procedures.

But Steve Lott, spokesman for the International Air Transport Association, which represents roughly 230 airlines around the world, says that there have been delays.

Heaven help us, right? It's the classic dilemma of any political battle. Facts become fungible. Whether the subject of dispute is the size of a political rally or the consequences of a new mining practice, control of data and presentation of statistics become tools of fear, uncertainty, and doubt. Being able to separate spin from substance in the digital world will be increasingly important in the 21st century. To do that, citizens need to be empowered with reasonably reliable data and the tools necessary to analyze the relevant chunks for themselves. Projects like Sulkin's are experiments in accomplishing that goal.

Of course it's unclear if FlyOnTime, or any amount of data, will be at all relevant or influential in this discussion. As with the Glenn Beck (and Jon Stewart/Steven Colbert) rally, lots of people showed up on the National Mall no matter how many there were, exactly. As with hydrofracking, some people can still ignite the water coming out of their hoses and kitchen sinks, no matter what the environmental reports say.

But there's an interesting moment about to happen here: A project that has been designed to provide insight on a hard-to-quantify question using a mix of government and citizen-generated data — one that, like many such projects, is a little hangdog in the traffic department — meeting a point in time where its subject matter dominates the news cycle.

"In a crowdsourcing experiment, you really need either to have crowdsourcing be incidental to some other task," Sulkin told me by phone Wednesday, "or provide some kind of incentive, [some] other way to get people to do it."

Both the incident and incentive, it seems, have arrived. The outcome of this experiment will be worth watching.