I am pretty sure you have already heard abou the Air Quality Egg: It is a device for monitoring the air quality and reporting results on online repositories like Pachube. It is based on the Nanode and comes with various sensors that measure the air quality in the immediate environment.
If you want to become part of the project and receive your own egg you can support it through Kickstarter (7 days left!)
Based on the number of sensors (and data) that will be available the AQE looks like a very interesting project and challenge for applying pattern recognition techniques in order to either identify patterns (based on location of eggs, sensors used, etc.) and/or classify air quality (based on user estimation, correlation with other data, etc.). Apart from collecting the data this project would need some additional information from users (such as air quality estimation, etc.) and a mechanism for analyzing the data. It could be implemented as part of the back-end infrastructure of Pachube or completely as an external application. The outcome could be used to provide an automated estimation of air quality, show what sensor data have the highest impact on air quality etc. It could also be used by the scientific community to perform further experiments in the domain since data will be somewhat structured and annotated with air quality information.
There is already an open discussion for this topic:
If you are interested in participating simply complete this survey.