Last week, it was announced that Chicago would be getting its own "fitness tracker", sometime in late July—a lamp post-mounted civic Fitbit that gathers data about the city's health and motion, both communicating it immediately through a simple interface and storing it for ongoing study.
It's an ambitious project, the brainchild of scientists at the Urban Center for Computational Data at the University of Chicago. And it has an ambitious name: the Array of Things.
But, like a fitness tracker, the measurements are pretty basic: temperature, humidity, light, sound, carbon monoxide, nitrogen dioxide, motion, low-resolution infrared, and a Bluetooth modem to (theoretically) count foot traffic. The idea is that anyone will be able to sample the immediate health of the city with a glance at the Array—current carbon monoxide or nitrogen dioxide levels, perhaps. Or a combination of temperature, humidity, and roadway temperature (detected by the infrared sensor) could warn drivers that a street or bridge is perilously close to icing over, provided the math works. And anyone will be able to gather that data as it collects to look at long-term trends.
The plan caused a bit of a stir, particularly for an open-data initiative, because of that last sensor, the Bluetooth modem, which will ping any devices in the immediate area, in a game of wireless Marco Polo. When the devices ping back to the Array node, it'll register that there's a person there.
Doing so a couple times a minute is supposed to give a measure of foot traffic, but the specter of surveillance drew a lot of attention: "Chicago Gets a New Surveillance System Straight Out of a Video Game"; "Big Brother? Chicago to measure pedestrians' movements"; "Chicago gets a friendly Big Brother"; "a new high-tech surveillance system promises to create utopia." Oh, and drone jokes. Ald. Fioretti wants hearings to address privacy concerns and because "officials could potentially have netted millions of dollars by reaching an agreement with a data collection agency."
What will the Bluetooth sensor actually do? "There's one command that sends a broadcast, and it asks for Bluetooth devices active in the area to respond," said Charlie Catlett, senior computer scientist at Argonne National Laboratory and director of the Urban CCD, in a phone interview this week. "In that response is the hardware address for that modem that responded…. In order to get data or put data onto a device, we have to go one step further, and then we'd have to respond with a request command. At that point the owner of the device would be given a prompt that says 'do you wish to connect to this other thing?' We're just taking step zero, which is to ping the devices, then we're throwing the addresses away and just counting the number of responses we got.
"You already have the situation as you're walking through the city that there are Bluetooth devices and WiFi devices around you," Catlett says. "The difference between what we're doing and someone might do in a coffee shop over the coffee shop WiFi, there isn't any visibility or transparency to what's happening in the coffee shop, whereas everything that we're doing—the software's published as open source, it's approved by the city, among other groups."
(There's also no guarantee that it'll actually work—not everyone has a Bluetooth phone, and not everyone who has one actually enables Bluetooth. Some places could have a comparatively high percentage of Bluetooth users, like LaSalle Street near the Board of Trade, I'd expect, so even adjusting for use could be difficult. It's an experiment; if it worked, it would be an efficient alternative to the typically laborious process of doing pedestrian counts. But it's a tricky problem.)
Not only will the data be open—through the city's Data Portal—but the software and hardware will be open-source as well, once complete. So watchdogs, not to mention developers, will be able to see what the Array is doing with the data. Further, the hardware will be open-source as well, meaning that people could print their own circuit boards, connect them to the (also open-source) Arduino microcontroller at the node's heart.
In theory, people could build their own nodes, in much the same way that, say, the popular weather site Weather Underground has some 30,000 users submitting data from personal weather stations. "We've been thinking about how could we set up so that people could not only set up their own device, which they could do, but they could somehow contribute their data to the Array of Things," Catlett says.
"There is a project that some of our collaborators have done in Barcelona called Smart Citizen. It's a Kickstarted project, and they do five sensors, and it's completely crowdsourced. Everyone who buys one of their devices connects to their database." In Amsterdam, for instance, they've been using Smart Citizen to record noise levels in the city, though Catlett cautions that opening up the Array would require careful attention to data-quality issues.
Mostly, the Array will be gathering environmental data. Aside from the aforementioned sensors, the Urban CCD want to eventually include precipitation, wind, carbon dioxide, and volatile organic compounds, which would give the city (and its citizens) real-time access to micro-level environmental conditions.
"The city already does this clever trick with the GPS tracking of buses, where they correctly observed that the movement of buses is a reasonably good proxy for the movement of the rest of traffic," Catlett says. "That's how they estimate congestion in the downtown area. If you combine that with emissions, you could get a sense of the impact of that traffic on air quality. Maybe you'll find out that the worst time to walk down a particular street for emissions isn't actually during rush hour, but it's during the hour after rush hour. If you wanted to navigate through the city based on air quality, you'd really want something that was on every street corner."
The Array of Things is still very much in flux, though. The Urban CCD's main priority is getting the nodes to collect and publish the data. They're planned to roll out slowly, beginning downtown: first, one at Jackson and Michigan at the end of July, then several running north to Wacker by the end of August. If the pilot program is approved and works, the next step would be 27 nodes within the Loop. If funding from the National Science Foundation comes through, it would spread to other neighborhoods, finding new uses along the way. One community's priority might be light and sound, where another could be interested in air quality.
Or the interaction between the Array, Chicagoans, and their city could be even more direct and complex. "There's a group at the University of Chicago, for example, that's doing interactive and augmented-reality games for and with high schoolers, especially on the South Side, to give them ways to rethink their environment, if you will," Catlett says. "We've talked to groups like that, where there's some interest in possibly delivering information services in a hyperlocal way."