Dr. Nipun Batra is an Assistant Professor at IIT-Gandinagar. He has a PhD. from IIIT Delhi and completed his postdoctoral research at the University of Virginia. His group broadly works on machine learning/artificial intelligence/sensors/IoT for computational sustainability problems like smart buildings and air quality. We recently spoke with Dr. Batra about his work.
Can you tell us a bit about your research?
I’m a computer scientist by training. My research broadly involves applying machine learning methods and developing sensors for sustainability applications. During my PhD, I worked on smart buildings. Specifically, I focused on an application called energy breakdown – providing a household with detailed per appliance energy usage using minimal sensing. Such an energy breakdown, like an itemised grocery bill can help occupants save energy.
What got you started on the topic of air quality?
I think there are several reasons behind working on air quality. First, I was born and raised in Delhi. Over the years, I have seen the air quality decline to the now hazardous levels. My father has been asthmatic and I remember him using an air purifier (especially around Diwali) before they were a common household thing. I have also seen young children from my neighbourhood diagnosed with respiratory disorders at a very young age. It sends a chill down the spine! I hope collectively as a nation (and world) we are able to take some concrete steps towards mitigation of pollution and I am trying to play my part. Second, from the technological standpoint, I identified various deep problems that I could work on as a computer scientist. Some of these were similar to the problems I worked during my PhD, and others while different, would be ones I’d love to work on. Third, I like to work on interdisciplinary problems and specifically those who have local importance. I have found the air quality community to be supportive and willing to collaborate.
Recently, you’ve been using data from social media to understand perceptions about air pollution in India. Can you tell us more about this work?
Social media can often tell the pulse of what is going on in the country. The scale is unprecedented. In our recently published work, we looked at Twitter data in connection with Delhi air quality and made some startling discoveries. First, there is a lot of support for unproven/untested/unscientific mitigation strategies like installation of smog towers. Such support pervaded different strata of the social media discussion barring a handful of researchers who understand that such strategies are infeasible. Second, we (unfortunately) found that the discussion on social media is very sparse and episodic. Yes, you guessed it, it is mostly concentrated around winter months, even though the pollution is clearly an year long problem. Even news media and popular figures mostly discuss the problem only around the winter when it is already too late! A lot of discussion also springs when people can “sense” poor air quality — either via the haze or smell; forgetting that the air is severely polluted throughout.
Can you tell us about your ongoing projects on air quality?
Yes, we are working towards various aspects of air quality. Much of the work is in its preliminary stages and I would be open to collaborations, joint proposals, consultancies. I’ll now describe five important projects beyond the social media air quality perception project I mentioned earlier.
- First, we have developed (and continue working on) a smart mask that can be used as a ubiquitous medium for low-cost air pollution sensing and lung function testing. The key idea being that a small microphone in your mask can be used to gauge your lung function. Such sensing can provide accurate exposure measurements along with real-time health effects.
- Second, we did some initial studies on the differences in pollution exposure of different people inside an academic campus. The numbers are alarming! The blue collar workers have a significantly higher exposure due to their nature of work despite being in overall similar ambient conditions to their white collar counterparts.
- Third, we are working with a Google grant supported project on identifying the role of air pollution in development of co or reinfections in COVID recovered patients. This project is in its infancy. The goal of the project is to identify the most vulnerable population and appropriately budget and intervene.
- Fourth, we are working towards development of novel interpretable machine learning algorithms for fine-grained spatial air quality prediction. We are also working on proposing algorithms to recommend locations for installing air quality sensors. Both these ideas are based on the fact that we have a significantly less than ideal number of air quality monitors in India. Installing each new sensor is significantly expensive in both capital and operational cost, and thus, we need a smart strategy to do so. Our initial studies suggest that being strategic in sensor placement is significantly better than placing them randomly.
- Fifth, we have done some research on developing tools for air quality visualization. The goal is to lower the barrier to entry via well documented open source tools.
Are you accepting new students?
Yes, I would be happy to work with students/junior research fellows (JRFs). I have a dedicated openings page where I ask the prospective students to work through a series of questions and implementations: https://nipunbatra.github.io/openings/
Given the interdisciplinary nature of my work, we have worked in machine learning, social networks, sensor networks, ubiquitous computing, among others. Interested people should contact me via my official email id: firstname.lastname@example.org
Learn more about Dr. Batra’s work on his website.