A pilot model of an artificial intelligence (AI)-based system designed to identify vehicle pollution hotspots in real time has been developed by CSIR-NEERI. Dr S Venkata Mohan, Director, NEERI; Dr K V George, Chief Scientist and Head, Air Quality Management Division; Dr Prakash Kumbhare, Senior Principal Scientist; Dr Rahul Vyawahare, Senior Scientist, shared the details of the innovation with mediapersons on Tuesday at NEERI. The real time monitoring of emissions would help the authorities understand the origins and enable policy decisions to tackle the issue. 

 

 

Dr Vyawahare, Dr George, along with Atharva Malode, Raj Sonarghare, Jay Singh Rajput, and Pravan Nair have developed an AI-Integrated Line Source Emission Inventory (LSEI) dashboard that can analyse live traffic footage and estimate vehicular emissions in real time. The system processes CCTV feeds from city roads and uses AI to identify and count different categories of vehicles such as two-wheelers, three-wheelers, cars, light-duty vehicles, heavy-duty vehicles and buses. Based on the number and type of vehicles detected, the dashboard calculates emissions such as particulate matter (PM), nitrogen oxide (NOx), carbon monoxide (CO) and hydrocarbons (HC). According to the scientists, the system can help authorities identify high-emission areas and understand peak pollution periods, enabling more effective traffic management and air quality planning. Officials added that once implemented on a larger scale, the system could provide evidence to fast-track decisions aimed at reducing vehicular pollution and managing traffic congestion in Nagpur. 

 

 

Officials said, the AI system has been trained extensively using traffic camera footage from multiple cities across the country, including Delhi, Kanpur, Jalgaon, Nashik, Akola, and others, enabling it to accurately recognise different categories of vehicles. However, at present, the system cannot reliably distinguish between electric and regular vehicles. Given that electric vehicles constitute a very small proportion in Nagpur, scientists are currently accounting for them through manual removal while training the AI to improve identification. 

 

 

At present, the system is operating through a camera installed outside Gate No 1 of the NEERI campus near Ajni Square on Wardha Road. The Institute is in discussions with the Nagpur Municipal Corporation (NMC) and the city Traffic Department to expand the deployment of the system to additional locations across the city. Researchers explained that conventional emission inventory often relies on compiled datasets and may involve feeding huge volumes of data to generate estimates. In contrast, the AI-based system produces real-time emission estimates by analysing actual traffic conditions.

 

 

The LSEI dashboard is not currently available for public access. It operates on a temporary IP address for internal use. Scientists stated that, the system will soon be moved to a permanent IP address so that everyone can access the information. As soon civic and other authorities show interest, CSIR-NEERI can scale up the model and deploy it for widespread use.