Presented by: Charul Paliwal, IIIT Delhi

Summary: Measuring the local air quality is a reliable proxy of measuring the emissions in a given location.  The measurement of air quality parameters in a locality is the first crucial step towards the localization of credible pollution sources like local industrial units, transportation, or even crop burning. Dense air quality monitoring in most cities/towns, especially in the developing countries, is virtually non-existent due to the expensive static sensors setup. 
The central theme of this work is on answering the following questions: 1) How can we develop a dense spatiotemporal air quality map in a city using only few monitors?  2) Does building a network of mobile monitors along with static monitors help in creating a dense air quality map?

We propose the creation of a dense air quality map by performing a spatiotemporal sampling using only a few moving sensors followed by the suggested matrix completion technique. The underlying low rank and slowly time-varying structure of the air quality data can be leveraged to create models that facilitate an effective spatiotemporal extrapolation.  These mobile monitors are put in the public transit buses as well as small vehicles to ensure dense coverge.  We claim that to obtain a dense air quality map; the cities may require only a small fraction of moving sensors as compared to all static sensors setup. Therefore moving air quality measurement could lead to an effective strategy against air pollution and climate change.

We are further left with research questions on : 1) What is the best moving sensor strategy? 2) Should we put a mixture of accurate (industry grade) sensors and low-cost monitors?  3) How to calibrate moving sensors using the industry-grade static sensors  4) What cost savings can be delivered with such a setup? 5) Additionally, how do we localize the cause of pollution using such a setup?

Click here for presentation video!