CDAC to pilot AI-driven precision farming project in Jharkhand

Kolkata, Jan 9 (PTI) The Centre for Development of Advanced Computing (CDAC) will pilot a project in Jharkhand to support micro and small farmers through advisories using an AI-driven precision agriculture model based on free satellite imagery and weather data, a senior scientist said here on Friday.

The pilot project, which has received initial approval from the Department of Science and Technology, aims to address key challenges such as water scarcity, soil degradation and inefficient use of farm inputs without requiring farmers to invest in costly physical sensors, CDAC scientist Alokesh Ghosh told PTI on the sidelines of "AI in Agriculture".

It was a pre-summit event ahead of the India AI Impact Summit 2026, scheduled to be held in February, which is being positioned as the first Global South summit on artificial intelligence and is expected to see participation from around 100 countries, according to the Ministry of Electronics and Information Technology.

Ghosh said the proposed system is a low-cost, cloud-based decision support platform that integrates satellite imagery, weather inputs from the India Meteorological Department (IMD), soil health information and field-level data to generate location-specific advisories.

"One of the biggest barriers for Indian farmers is the high cost of sensors. In the Jharkhand pilot, we have eliminated the need for field hardware by using satellite and weather data to estimate crop water and nutrient requirements," he said.

A key feature of the Jharkhand project is micro-level weather forecasting at the panchayat level, with a grid resolution of 4x4 or 5x5 square kilometres, allowing highly localised, crop-specific advisories.

"The information is now available at a micro-level across India, but for Jharkhand, we have specifically created this grid system for the pilot," Ghosh said, adding that farmers growing different crops in the same panchayat area receive separate advisories on their mobile phones.

The system requires farmers to geo-tag their plot boundaries and provide basic details such as crop type and sowing dates. AI and machine learning models then generate short-, medium- and long-term weather-linked advisories on irrigation, fertiliser and nutrient management through digital platforms and SMS.

The project also seeks to reduce water wastage in agriculture, which accounts for nearly 90 per cent of total water use in India, much of it lost due to inefficient irrigation practices, said Prof Amlan Chakraborti of the University of Calcutta's School of Information Technology.

By providing precise irrigation and nutrient recommendations, the model aims to improve water and fertiliser use efficiency while preventing soil stress caused by excessive input application, said Prof Debashish Mazumdar, senior professor at Sister Nivedita University and former director.

While currently limited to Jharkhand, Ghosh said the pilot could serve as a scalable and affordable precision agriculture framework for wider deployment across the country, particularly for managing hybrid crops that require targeted inputs.

"The objective is to optimise yield from limited cultivable land while conserving water resources," he said. PTI BSM NN