Seema Chavan, 56, a farmer from Swami Chincholi village in Baramati Taluka, in Maharashtra’s Pune district, has suddenly earned worldwide fame. In a video clip that Microsoft CEO, Satya Nadella, shared twice on X, Seema talks about her experience of using AI in her sugarcane farming. “My cost of farming has come down by at least 30 per cent and the yield is up by 30 to 40 per cent. I use AI on 1.5 acres of my farm. But from next year, I will bring my entire field under the AI,” she said reflecting on what the future of farming could look like, aided by AI.
The world has already begun discussing AI’s benefits to farming. But thus far, AI’s use in the actual growth of crops has not been debated. And the Baramati experiment appears to be pioneering this prospect.
The experiment began three years ago, at Baramati’s Agricultural Development Trust (ADT), a reputed institution established in 1971 by MP Sharad Pawar and his elder brother Appasaheb Pawar. When Sharad Pawar and his younger brother Pratap Pawar started reading about possible changes that AI could bring about in people’s lives and livelihoods, they became curious about the possible use of technology in agriculture. ADT was already in a tie-up with Oxford University for knowledge-building in farming. So ADT invited Ajit Jaokar, an engineering scientist and course director for several AI programmes,to visit ADT and Krishi Vigyan Kendra (KVK) of Baramati.
Also Read | Gains for sugar cane growers in Maharashtra
Jaokar, pleased with ADT’s scientific approach to farming, introduced the brothers to Ranvir Chandra, the Chief Technology Officerof the agri-food section ofMicrosoft, who promised technical support from Microsoft for research. It was a coup for ADT, with Microsoft and Oxford on board. Before long, ADT in Baramati began working towards creating an AI centre at its campus.
One crop, two ways
Now it was time to experiment with the new technology, and ADT began to do so on their own farms, ranging from sugarcane to tomato. To assess the benefits of AI, they grew the same variety of crops in two ways: one assisted by AI and the second used traditional farming methods. The result was startling: the crop grown with AI assistance fared far better in terms of yield. “Farmers who visited began to notice the difference; this garnered their support, and we were able to expand the scope of the experiment,” said Prataprao Pawar.
A 24×7 war room at KVK, Baramati, provides real-time support to farmers using AI-driven insights.
| Photo Credit:
By Special Arrangement
Last year, around 1,000 farmers expressed interest in adopting AI in their farms, especially for sugarcane. So, 1,000 farmers from across Maharashtra, where sugarcane is grown, joined the AI farming experiment. And Seema Chavan was one of them.
ADT went about their project methodically. First, a “war room” was set up in KVK. Representatives of Fasal (a start-up that helps farmers get information about weather and soil quality), Map My Crop (which uses satellite imagery to map farm areas) and Agri Pilot (which processes data from weather stations, soil sensors and satellite maps) were asked to be available 24/7 at the room.
ADT designed a standard operating procedure with the help of experts. First, they had to physically reach farms to collect soil samples and longitude-latitude data; weather stations, and soil sensors designed by Fasal were set up on every farm; and Map My Crop mapped the farm area with satellites. Meanwhile, Oxford researchers analysed the data and created an algorithm that was given to Microsoft’s AI system, which processed the data, and included traditional knowledge fed by KVK.
KVK collected data directly from farmers and this was compared with AI’s suggestions. Yogesh Phatake, a microbiologist with KVK said, “We found that almost 95 per cent of AI findings and ground information matched. This helped us to garner farmers’ confidence to continue with the project.”
This was just the beginning of raising a crop. All 1,000 farmers were asked to download an app, Krushik, designed by KVK. The information gathered and processed by AI was transmitted to farmers through this single window. So, weather station information could be seen by the farmers instantly, as was data from soil sensors. AI could now even send farmers forecast alerts.
Ajit Jaokar (right) with Sharad Pawar. Jaokar is an engineering scientist and AI expert who directs several AI programmes.
| Photo Credit:
By Special Arrangement
Vilas Bhagat, a progressive farmer from Korhale village in Baramati was initially sceptical about AI. But his son Aditya, encouraged his family to enrol for the experiment. The Bhagat family, with 140 acres of sugarcane, decided to try out the experiment on two acres. One day, Aditya got an alert that there would be 3 mm of rain at his farm one afternoon. He was at his home and the sky was clear; members of his family laughed at him. But he went to assess the situation at the farm, there was indeed a shower. “This was when my family started taking the project seriously,” said Aditya.
Pest alert
Another advantage was disease prediction through satellite data. One day, Aditya got an alert that a part of his sugarcane crop was at risk of trunk worm infestation. When Aditya and his father Vilas went to check they found that the satellite information was accurate. “We immediately sprayed preventative medicine. We in fact sprayed the entire farm to save the crop from disease,” said Vilas. “One alert saved us almost Rs.80,000,” said Vilas Bhagat.
Soil sensors also collect real-time data on moisture and temperature, which is processed with the help of AI that indicates the exact timing of irrigation. Seema Chavan said, “Earlier I used to irrigate the farm for six to seven hours twice a week. But this time has now reduced. I learnt that watering the farm for two hours is enough.
The timing and quantity of fertilizers needed are also provided through AI. “AI also helps us with the exact location of fertiliser requirements. This has reduced the quantity of fertilizer we use,” said Aditya Bhagat. “My expenses on fertilizer, water, as well as labour, has reduced by 30 per cent,” said Seema Chavan.
Ranvir Chandra, the Chief Technology Officer of Microsoft’s Agri-Food division, visits the site. Chandra specialises in AI-driven solutions for agriculture. He played a key role in bringing Microsoft’s technical expertise to the AI-powered farming experiment in Baramati
| Photo Credit:
By Special Arrangement
Meanwhile, the farmers have noticed another promising outcome of AI-assisted farming: while traditional sugarcane grows is 8 feet tall, crops monitored by AI are around 10 to 11 feet tall; and the sucrose percentage, which ultimately decides the farmers’ income, has increased by 2 to 4 per cent, depends on the variety of the plant.
With the success of their experiment, ADT Baramati is now expanding their project there are plans to experiment on one lakh ha of farmland, involving 50,000 farmers. Online registration has reached 12,000 now, but offline registrations would be double that. The Maharashtra government in its budget presented on March 10 announced a Rs.500 crore financial assistance package for experiments in AI in agriculture.
ADT, Baramati, is already working on the issue. The State’s Finance Minister, Ajit Pawar, the now estranged nephew of Sharad Pawar, made this announcement. “We are sure we will easily reach 50,000 farmers because in last two months, farmers across States such as Uttar Pradesh and Karnataka have come to see the experiment,” said Nilesh Nalawade, CEO of ADT, Baramati. “But right now, our focus is on Maharashtra.”
Also Read | Bitter aftertaste
However, such technology-based programmes that create huge amounts of data must be protected by laws on its usage, and the government has a role to play in assuring farmers about the security of the information gathered from the farms.
But there is another big challenge ahead: To make the technology cost-effective in the Indian farming context. KVK’s weather machines, sensors, etc. were provided at a highly subsidised cost (Rs.12,000) and marginalised farmers can’t spend the actual amount (Rs.80,000) for the new technology. So, it is now up to the government to take the baton forward.
While visiting the ADT, Baramati in January 2023, Chandra termed the experiment “Farm of the Future”. These words could well be prophetic, not just in Baramati, but for the country as a whole.
