Today's sophisticated sensors can map a whole city, block by block, detecting pollution that often varies by street.
EDF is bringing together innovators from industry, universities and advocacy groups to unlock this opportunity.
We're making once-invisible health and environmental hazards visible.
In one of our projects, Google Street View cars equipped with sensors generated a trove of pollution data that can lead to custom solutions in West Oakland, California.
Here are the challenges ahead
Through our work, we've seen great potential for sensors as a low-cost tool for solving environmental problems. That reality is closer than ever. Addressing the roadblocks, such as those below, is key.
1. Managing sensor data is complex
Air pollution is made up of different pollutants – such as black carbon, ozone, nitric oxide and nitrogen dioxide – that cause health problems.
It's important to track all to get a full picture of this often invisible threat.
Different pollutants require different types of sensors, calibrated at different schedules.
Sensors that measure nitrogen oxides and other air pollutants, for example, often must be calibrated more frequently than those that measure harmful particle pollution like soot.
Plus, how and where you measure matters. To get a fine-grained data set from sensors mounted to a moving vehicle, you need to collect data as often as every second. For stationary monitors, sampling hourly may suffice.
2. Low-cost sensors need refining
Traditional air sensors are sophisticated scientific-grade equipment designed to sit still and get regular service.
Low-cost sensors are less durable. Depending on how they're used, parts can break down and have to be replaced.
To expand projects like the one in West Oakland, we need to continue to deploy the latest technology in pilots around the globe to test these sensors in real-world situations.
3. Data analysis raises questions
EDF has convened a group of scientists to determine how to collect the right mix of data in the most efficient way.
They are deploying mobile sensors in several U.S. cities – along with some higher-cost stationary sensors still needed to validate data from the less-expensive mobile devices.
The West Oakland project showed how to map stable spatial trends in air pollution – with researchers devising one way to handle the sheer size of the data set.
Some questions they encountered: How do you interpret one-second data, and should you take the average over 10 seconds to get 10-second data? And how do you roll that into an annual average?
To accurately answer these questions we've had to secure computing resources and request extensive feedback from peer scientists. Until data collections of this scale become routine, this process continues each time we map a new city where conditions are different.
4. Vastly different players must work together
Our role at EDF is to develop and catalyze a broad network of experts and advocates to make sure the maps are produced in a scientifically sound and transparent way – and to ensure they provide maximum benefit to society.
Entrepreneurs, for example, are necessary to develop and bring new high-quality, low-cost sensors to market.
State and local agencies can share existing data and analyze new data, and then set policy to reduce air pollution.
Community organizations, meanwhile, can provide critical local knowledge that will help researchers to collect good data and help define the analysis necessary to advocate for policy solutions that will cut air pollution in their neighborhoods.
Such an undertaking requires vision, extraordinary commitment and investment over time. We're on our way.