Inspections are central to ensuring facilities comply with environmental laws and regulations. The Maharashtra Pollution Control Board (MPCB) will be the first outside the United States to pilot the use of a new machine learning-based artificial intelligence (AI) system built to help the state identify which facilities regulated under the Air Act to inspect, with the help of researchers at the University of Chicago. MPCB and the University of Chicago Trust, through EPIC India, agreed to work together to advance this effort.
An MoU to advance this initiative was formalised at an event in Mumbai on 17 February 2026, in the presence of Smt. Pankaja Munde, Minister of Environment and Climate Change, Government of Maharashtra, who attended as the Chief Guest. The MoU was exchanged between Shri Dinesh Sonawane, Statistical Officer, MPCB, and Ms. Nidhi Sarin, Executive Director, India Program, Institute for Climate and Sustainable Growth, University of Chicago. Other special guests included Shri Siddhesh Kadam, Chairperson, MPCB; Shri Balaji Srinivasan, Co-Chair, Emissions Market Accelerator (EMA); and Ms. Smita Rakesh, Director, Earthshot India.
“As we become increasingly urbanised, with more technology and industries, we are also experiencing the growing challenges of pollution. We need to look at how technology, including AI, can help us monitor industrial pollution more intelligently and identify ways to make our systems more effective.
Our role is not just to regulate industries or control pollution. We must also work alongside industry and encourage participation to create solutions that benefit both the environment and economic growth. We need to think beyond the immediate challenges and work towards leaving behind a cleaner, healthier environment for our future generations,” says Smt Pankaja Munde, Minister of Environment and Climate Change, Government of Maharashtra.
The new AI system was built by researchers at the University of Chicago’s Energy & Environment Lab, who put it to the test in partnership with the U.S. Environmental Protection Agency (U.S. EPA). Using 15 years of historical data, the researchers developed a model to predict which hazardous waste facilities were most likely to have serious violations. When they compared the AI-based inspection targets with traditional decision-making in regulatory inspections they found that the model detected 82 percent more violations than the traditional approach. The U.S. EPA is now using the model to inform site selection for hazardous waste facility inspections nationwide.
“No regulator has enough inspectors. That’s just the reality. What AI does is help them make better choices with what they have—go where the risk is highest, not just where it’s convenient,” says Michael Greenstone, Milton Friedman Distinguished Service Professor in Economics and director of EPIC. “In a state like Maharashtra, with tens of thousands of facilities, this can be a gamechanger.”
While the model is being used in the United States, the same approach is now being adapted to a new type of pollution for the first time in Maharashtra. The approach will be applied to a subset of industrial facilities in Maharashtra, one of India’s most industrialized states. If the partnership is successful, the pilot could expand to all the approximate 53,000 facilities regulated by the MPCB.
“Our work with the U.S. EPA showed that our machine learning approach can help regulators target inspections more effectively,” says Olga Rostapshova, executive director of the Lab. “We are excited to build on that success through this project with the state of Maharashtra that has pioneered efforts to improve air quality in India and beyond.”