CitiSense – A Participatory Air Quality Sensing System for Real-Time User Feedback

Recent revelations about the impacts of air pollution on our health are troubling, yet air pollution and the risks it poses to us are largely invisible. Today, the infrastructure of our regulatory institutions is inadequate for the cause: sensors are few, often far from where we live, and the results are slow to come to us. What about the air quality on your jogging route or commute? Can you be told when it matters most? Advances in computing technologies can allow us to answer these questions. By pervasively monitoring ourselves and our immediate environs, aggregating the data for analysis, and reflecting the results back to us quickly, we can avoid toxic locales, appreciate the consequences of our individual behaviors, and together seek a mandate for change.

The CitiSense project is leveraging the proliferation of personal mobile computing via mobile phones and the adent of cheap, small sensors is developing a new kind of "citizen infrastructure". Challenges abound in power management, data security, privacy, intergerence with "noisy" commodity sensors, and incisive yet considerate user notification. An overriding challenge lies in the integration of the parts into a seamless yet module whole that can make the most of each component at every point in time through dynamic adaptation. Solving this probelm will not only allow the superior integration of existing techniques, but allow developing new adaptive techniques not before possible. In CitiSense we are investing the use of the Open Rich Services (ORS) publish-subscribe architecture to address these challenges. As just one example, ORS will enable highly adaptive power management that not only adapts to current device conditions, but also the nature of the data, the data's application, and the presense and status of other sensors in the area.

We have completed our second-generation prototype of the CitiSense system (See Figure). It comprises (a) a sensor board with Bluetooth networking that hosts nitrogen dioxide, carbon monoxide, and ozone sensors, (b) a back-end server that warehouses captured data and performs machine learning to interpolate the raw data to other locations and time, and (c) a phone application that connects to the board, relays sensor data to the back-end server, and displays information to the user.

With this system we are now achieving major results, including: a user study that shows how people incorporate real-time pollution data into their daily activities, results showing that such data holds significant promise for population health scientists, and a new power management result that uses the prediction models generated on the server to decide when to send new sensor data back to the server (when it contradicts the model).

Award ID: 0932403