It started with a Google search. When confronted with the dead animals washing up on the Gulf Coast after the BP oil spill in late 2010, cartographer Jeffrey Warren and a small community of technologists and volunteers typed in "how do you identify oil."
Warren and his team had been down in the Gulf documenting aerial views of the spill. While taking the imagery wasn't too hard—they hitched cameras to cheap kites and balloons and then open sourced their designs to the public—actually counting the number of animals killed by the spill was a much more complicated process. While BP continued to revise estimates of the size of the leak, environmental groups were waiting months upon months for the results of expensive lab tests to prove the spill was much bigger than BP admitted.
Warren and his fellow volunteers decided to take a shortcut. Largely using information found online with input from formally trained scientists, the volunteers built chemical identification tools from scratch. Their open source, DIY spectrometers made from hardware store materials could show, in basic ways, if the oil was to blame for the black gunk oozing from seagulls' beaks. Now, these devices can be attached to smartphones.
Three years after the spill, Warren and his colleagues, now a nonprofit called the Public Laboratory for Open Technology and Science, is spreading open-source pollution identification tools to as many non-scientists as possible. In January, the Lab initiative, the "Homebrew Sensing Project," won a $350,000 Knight Foundation health data award for its vision of putting the power to detect and identify pollution in the hands of communities all over the world. Public Labs is also building an online library of its samples to make it easy to contrast and compare examples of contamination.
"We had come up with this new way to collect evidence, and to collect data about pollution. And it was pretty cheap. You could do it by using materials around your home, or that you could by at Home Depot," Warren says.
Public Lab's new tools are now more advanced than in the early days, when volunteers would simply shine laser pens through mason jars to see if petroleum in the sample would glow. One of the Homebrew Sensing Project's infrared spectrometers can be folded out of paper and attached to a smartphone camera. Another is a spectrometer that points to Gulf refinery smokestacks and analyzes the pollution spewed from them into the air. Public Lab spectrometers have now been used to gauge yeast levels in beer and measure harmful elements in household cleaners. And Public Lab plans on expanding its repertoire.
Warren also suggests that the Public Lab approach to generating and analyzing these health and pollution data sets is a counterpoint to the trend of "big data." Instead of having faraway experts gather, analyze, and sell data, most often unknowingly on behalf of the people who generate it, Public Lab wants the citizen scientists who track the environmental health of their neighborhoods to be involved with every step of the process. He calls his way of thinking "small data."
"The people who use [the data], who are going to analyze it, and who will participate in the debates that the data generates—they're the people collecting it from their own backyards," Warren says. "They're who the data is about."