High demand for food quality and quantity has sparked a change in agriculture. Various research reports state that by 2050, globally, we will need to feed 2 billion more people without an increase of cultivatable land. They also say that over the next 40 years, the agriculture sector will need to produce more food than the industry has grown during the last 8,000 years. Thus, increasing the productivity from the existing arable land is the solution for rising food demand.  

While various agri-tech solutions are available currently to increase the yield, many barriers have prevented farmers in adopting digital transformation. For example, too much manual work to make the solution function or rely on remote internet accessibility are some of the major challenges. However, tremendous amounts of agricultural data are generated but never used. Technology company IBM, has come up with a much easier solution to overcome these obstacles with the help of AI and predictive analytics. IBM’s Watson Decision Platform offers hyper-local, (a village-level or even a farm level), weather information including weather forecast and soil moisture information, helping farmers to make decisions on water and crop management. 

“It is important for a farmer to know when it will rain. If AI can predict the rain accurately and how much and when that will happen, farmers can make meaningful decisions,” says Himanshu Goyal, India Business Leader, The Weather Company, IBM. “Watson platform will provide these type of accurate data so that farmers can decide when to irrigate, add fertilisers or delay some decisions,” he says. Weather data will also predict diseases. 

AI solutions

IBM’s solution provides information on productivity assessments, decision guidance and probabilistic weather conditions that feature a detailed analysis of sub-seasonal and seasonal forecasts. “The precision agriculture solution combines AI and weather data to obtain and analyse farm-level insights,” Goyal says. The Watson Decision Platform for agriculture applies AI, machine learning, and advanced analytics to the farm data to extract valuable insights and automatically generate guidance for smarter decisions. A unified dashboard enables farmers to visualise data and alerts related to critical elements such as weather forecasts, soil conditions, evapotranspiration rates, and crop stress. “For instance, AI visual recognition of drone-captured footage may be used to automatically identify certain types and severity levels of pest and disease damage,” he says. Farmers could save time, money and also reduce the impact of pests if they understood how and when to spray pesticides. The AI data also provides insights around irrigation, product application, and planting and harvest timing.

Projects with government

IBM has partnered with many government institutions and private companies and has helped a lot of farmers across the country to increase their farm productivity.  

Last year the company partnered with Ministry of Agriculture and Farmers Welfare to deploy this solution on a pilot basis for the Kharif crop season in three districts - Bhopal, Rajkot and Nanded - in Madhya Pradesh, Gujarat and Maharashtra, respectively. With its success, the project is now extended to ten districts. 

Another recent project is with Karnataka. The state implemented IBM’s Watson platform to predict the prices of tomatoes and maise. IBM developed a price forecasting system for Karnataka Agricultural Prices Commission that predicts the market price trends for at least a fortnight and the production pattern of tomatoes. The dashboard uses IBM’s Watson Decision Platform and blends data from satellite imagery and weather data to assess the acreage and monitor crop health on a real-time basis. It can detect pest and disease infestations, estimate the tomato output and yield, and also forecast prices. Previously, the output estimates are based mainly on acreage data. Other data like the prices in major markets of neighbouring states, are also considered for the price forecast. “The solution was launched for the three major tomato- growing districts of Kolar, Chikkaballapur and Belgavi and two key maise-producing districts of Davangere and Haveri,” says Goyal.  

With more hyper-local data inputs, IBM is now focusing on finetuning the weather prediction. “We are focusing to ensure that the weather data and the soil factor data are perfected for most of the regions,” he sums up.