CEEW study: AI can cut Delhi air pollution, needs basics right

The Air Quality Early Warning System (AQEWS) can predict severe air quality index days with over 80% accuracy, according to a study by the Council on Energy, Environment and Water (CEEW). The system was developed after crises including the 2016 Great Smog and 2018 dust storms. Seven cities beyond Delhi now have their own AQEWS: Pune, Bengaluru, Hyderabad, Ahmedabad, Kolkata, Mumbai, and Jaipur.

CEEW argues that adopting AI-driven air quality decision support systems (AQDSS) must include end-user feedback from the start. A new application, VayuChat, developed by IIT Gandhinagar and CEEW, integrates data from over 300 low-cost sensors at construction sites in Thane. The Brihanmumbai Municipal Corporation has a similar collaboration with IIT Kanpur using 75 sensors.

The study calls for building capacity to use AI systems well and keeping humans in the loop to judge AI output. It also recommends developing secure, offline local AI models and notes the Centre's IndiaAI Mission as an encouraging step. The CAQM will use AQEWS forecasts to decide GRAP stages this winter.