Towards farms of the future: How AI is revolutionising agriculture in India

Technological interventions - on the part of the government and tech majors - are likely to lead to better agricultural practices, yields, and an improvement in the lives of farmers

Ask any tomato farmer in Kolar, Chikkaballapur or Belagavi districts of Karnataka what his biggest worry is, and his answer will be the price fluctuations of the crop. Thanks to unstable prices, farmers are almost never able to plan a definite production pattern for tomatoes. What makes matters worse is that whether there is a rise in prices or a slump, middlemen profit either way. And farmers are often deceived into selling cheap even when the demand for the crop is high in a neighbouring market.    

This is set to change. In a path-breaking initiative, the Karnataka Agricultural Price Commission has got technology major IBM to develop a smart platform that will use artificial intelligence (AI), big data and machine learning to predict the price patterns of the tomato crop. Launched this month, the platform will help farmers in these three districts take decisions on when and where to sell their crop and at what price. 

“It’s a dashboard that will predict the market price trends on any given day for the next fortnight,” says Sriram Raghavan, vice-president, IBM Research AI. Incidentally, this is India’s first market trend prediction tool for farmers. 

The dashboard will leverage IBM’s Watson Decision Platform for Agriculture apart from using AI, big data and machine learning. Data analytics will take into account historical data points for demand and supply. The price of the crop in major markets of neighbouring states will also be factored in. These price trends, shown in a band of 15 days on any given day, will be displayed on a dashboard at Karnataka’s Agricultural Produce Market Committee yards in the talukas, helping farmers take wise selling decisions.

Price forecasts will also be provided for maize in the districts of Davanagere and Haveri in Karnataka. 

But the first step of the process is to predict the output and yield of the crop. For this, satellite imagery will collect data on key factors such as the nutrition, moisture and water level of the soil and the acreage under the crop. Moreover, IBM’s weather platform, The Weather Company, will provide crucial weather data from the growing regions.

“The condition of the food crop is monitored through internet of things-enabled sensors. The collected data is then analysed through artificial intelligence,” says Raghavan. Sensors will also monitor the crop for pest and disease infestation, following which, interventions will be prescribed. 

It is from all these data points that the platform will predict the output of the crop. This is in contrast to the current system of output estimation that relies chiefly on acreage.

While the price forecast of crops is a new development, crop yield prediction has been in existence for some time. In May 2018, Niti Aayog had asked IBM to develop a crop yield prediction model. “Currently in operation in 10 districts across Assam, Bihar, Jharkhand, Madhya Pradesh, Maharashtra, Rajasthan and Uttar Pradesh, IBM is using AI to develop technological models for improving agricultural output,” says Raghavan.

Technology is helping farmers in others ways too. Chennai’s M S Swaminathan Research Foundation (MSSRF), in partnership with Oracle, is giving weather-based pest and disease forewarning information to farmers in three villages in the Nagapattinam district of Tamil Nadu.

The data from weather stations at these three villages are collated with the 30-year historical daily weather data of the region collected from the Pandit Jawaharlal Nehru College of Agriculture & Research Institute (PAJANCOA & RI) in Karaikal. This data are processed jointly by PAJANCOA & RI and the Tamil Nadu Agricultural University, Coimbatore, and Oracle to run the forewarning model.

“The project is currently in a pilot stage, but we have already seen that the the advisory on pest and pesticides have helped farmers save costs,” says R Nagarajan, head of GIS and Remote Sensing at MSSRF.

Technology-based applications are helping farmers take crucial decisions on sowing as well. The Hyderabad-based International Crops Research Institute for the Semi-Arid Tropics, in collaboration with Microsoft, has developed a sowing app that tells farmers in Andhra Pradesh and Telangana when to prepare the field, when to sow a particular crop or even what to sow. The idea is that these advisories will help farmers achieve optimal harvests. The app, which uses Microsoft’s Cortana Intelligence Suite, uses an interface between AI, machine learning, weather forecasting models and weather and agricultural data over several decades for the region. 

Innovations by agritech firms are also contributing to tech’s advance in agriculture. For example, CropIn Technology Solutions, a Bengaluru-based agritech firm, has built an app that gives information from the first stage of mapping of the farmland till the time of the harvest. “We are working on a smart agriculture project called ‘Jeevika’ in Bihar and Madhya Pradesh, where CropIn is helping farmers adopt better practices,” says Krishna Kumar, founder and CEO at CropIn.  

“We geotag the farmland and then collect data such as the health of the soil and water. The data are fed into our app, which triggers the package of practices for that plot,” Kumar says. 

Once the crop is sown, the app generates a report every five days on matters such as which part of the farm is not doing well and requires attention. Kumar says the CropIn app is backed by AI and machine learning. The firm also uses data analytics to predict crop outputs.

All these technological interventions — on the part of the government, tech majors as well as start-ups — are likely to lead to better agricultural practices, yields, and an improvement in the lives of farmers. But whether or not that actually takes place will depend on whether there are sustainable models with the scope for scaling up. Until that happens, the Indian farmer will remain at the mercy of the vagaries of the market and the rapaciousness of the middlemen who profit at their expense.