When it comes to the field of agriculture in the corporate world, S. Sivakumar is a name that seldom needs a long introduction. As the current Group Head Agri and IT Businesses, ITC Limited, Sivakumar oversees both agri as well as the IT Business of the Indian multinational conglomerate. Sivakumar currently is also the Chairman of Technico Agri Sciences Limited and Vice Chairman of ITC Infotech India Limited and its subsidiaries in the UK and USA. 

With a remarkable career spanning over 30 years, Sivakumar’s voice is one of the most valued and listened to when it comes to policymaking as well as corporate governance as evident by his role on the Advisory Council to the Ministry of Rural Development on National Rural Livelihoods Mission, the Commodity Derivatives Advisory Committee of SEBI and on the Management Committee of National Agricultural Higher Education Project of the Ministry of Agriculture & Farmers Welfare. 

In order to understand the impact of tools such as AI is having on agriculture and to gain a perspective on its future impact on country’s agriculture sector as well as the overall economy, we reached out to Mr Sivakumar.

How do you see emerging technology such as AI transforming the agriculture sector in our country? 

For times immemorial, farmers used their past experience of how crops responded to different agronomy practices and successfully dealt with the changing external conditions, such as weather or pest attacks. As the range of crops being grown multiplied with shifting consumer demand, and as the external conditions became more erratic due to the impact of global warming, the complexity involved in such decision making increased exponentially for the farmers. Depleting natural resources like water added to the challenge, leaving no option but to farm with more precision, unlike the past when such resources were abundant. 

If you have read between the lines and understood that a farmer is essentially “using memory and improving responses to different stimuli in a complex environment”, you may have already realised that agriculture is a fit case for AI. 

If we can integrate the hundreds of data points farmers produce on the ground with relevant external data sourced from a range of other stakeholders, and develop algorithms that can deliver decision support back to them real-time, AI would revolutionise agriculture. Enabling our farmers thus, and transforming them into knowledge workers, is a key step in accomplishing India’s mission of doubling farmer incomes.  

Which are the major domains in agriculture in which ITC use AI technologies?

ITC e-Choupal 4.0 offers a bouquet of services to the farmers like delivering information, co-creating knowledge, and assisting them with shared use of assets. At the core of this platform, custom-built by ITC Infotech, are digital crop monitoring & advisory, an e-marketplace for various farm inputs & equipment, and mechanism for selling farm produce. ITC works with several start-ups who plug into the platform and offer their point solutions to the farmers, stitched together by e-Choupal 4.0 as a meta market for the farmer. Such solutions include a hyper-local weather forecasting solution to align farming operations; field sensor-based precision irrigation initiative to improve the water use efficiency; remote sensing application for prediction of pest attacks to deliver early warnings; pesticide sprays using drones; IoT enabled post-harvest management practices inside hot curing chambers and cold stores for improving crop quality; image processing for rapid quality assaying at farmer doorstep etc. While some of these solutions already operate at scale, some are still being pilot tested for refinement.

We are also evaluating AI solutions to monitor crop and soil health to detect diseases and nutrient deficiencies. Autonomous robots to remove weeds when the field conditions are risky is another application being examined.

At the downstream end of the value chain, AI models are deployed in predicting commodity prices and optimising supply chain efficiencies.

And what is the impact so far? Is it lucrative to use AI and allied technological solutions?

AI is definitely helping agricultural operations to become more scientifically managed activities, with the ability to assess input needs more precisely and predict the output. As one illustration, ITC’s e-Choupal 4.0 initiative resulted in a 13% increase in crop productivity and a 27% rise in net returns per acre for the chilli farmers in Andhra Pradesh. The smart irrigation initiative of ITC has reduced the water consumption by 30% and achieved productivity improvement of 12% in tobacco crop. The price forecasts for different varieties of wheat were between 95 and 98% accuracy, while the yield forecasts at 92%.

Early successes are visible in hyperlocal weather forecasts and rapid quality assessment. The latter will transform how the transactions of agricultural produce are conducted along the value chain, especially at the farm end.

How lucrative depends on the scale of deployment and the value capture mechanism in the business model. 

Can you give some more details on how ITC is leveraging various technological solutions in establishing as a foremost leader in agribusiness?

ITC has been proactively harnessing various technologies by building as well as collaborating with the start-ups to address the pressing farmer problems across the value chain, viz. nursery, pre-harvest & post-harvest segments. Given below are the various technology interventions undertaken:

• Smart Irrigation: An IoT based automated solution for irrigation which controls water based on the crop needs in real-time. The technology includes remote sensing through satellites and on-ground soil moisture sensors to determine the soil moisture in real-time and communicating the same to the cloud through a LORA based network. 
• Weather Forecast: Hyper-local forecast providing accurate and targeted weather alerts to the farmers and translating them into actionable insights
• Satellite Use Cases: Leveraging satellite imagery with Machine learning for providing customised crop advisory, sowing and harvesting intelligence and monitoring the crop health for farmers. The solution employs a decision intelligence framework using multiple datasets which include satellite imagery, weather data, agronomic data, farmer profile data. 
• Drone Use Cases: Drone-based crop monitoring for forecasting any pest outbreak/disease incidence & providing actionable alerts to the farmers. It involves capturing the field images on a periodic basis with a hyperspectral camera mounted on the drones.
• e Kisanmitra: An m-commerce tool connecting farmers with agri input providers. The solution offers farmers reach to multiple suppliers with a wide range of inputs giving them an option to buy at an economical price.
• Farm First – Barn Monitoring System: An IoT based barn monitoring solution designed to monitor the temperature and RH in real-time inside the barn during curing operations and provide actionable alerts to curers. The solution is set up on a LoRaWAN network for data transmission in poor network areas. 
• e-Choupal 4.0 is ITC’s flagship initiative that carries all these solutions on a common digital platform. It also brings the communities together to discuss good farming practices and share farm equipment and other services. 

How do you build AI capability by crop? Is there a difference in approach when it comes to using AI and technologies like IoT for cash crops over staple crops?

It goes without saying that data is the foundation on which AI capability is built. The models need to be trained crop-wise data and need to be refreshed season after season with more and more data getting added to improve the accuracy levels.

Prima-facie, cash crops provide a higher scope of deploying technology-based solutions, as the unit value of these crops is much higher compared to staples, while the typical farmer-level acreages are also small. That’s why there are a number of interventions already being tried out in horticulture (fruits & vegetables). Costs of cultivation and risks are also high in these crops, offering more use-cases. On the other hand, because staples (rice, wheat) are grown on large acreages, there is an opportunity for deploying technologies that can make a difference at scale, e.g. satellite imagery, mechanisation solutions etc. in these crops.

What are the challenges in deploying AI solutions in agriculture, and how to overcome them?

While the potential benefits of AI solutions in agriculture are convincing, given that these technologies are still in the nascent stages of development, the cost of deployment is high. Farmers wouldn’t come forward to buy the services until the benefits are proven. Many enterprises are still struggling to figure out the value capture mechanisms, to be able to invest upfront. Because AI is not an installed product, but evolves over time and improves with usage data, this chicken & egg problem needs to be solved for taking the solutions to material scale. 

One key step is for the country to build an open, integrating framework that enables collaboration and interoperability across the Agri ecosystem. Another step is for the Government to build a foundational digital agri-stack with data collated in the first instance from farmer profiles, land records, soil health cards etc. that are made available to all the App Services. More aggregators like ITC e-Choupal will have to emerge with clear business models to carry all the point solution providers. These aggregators will need to work through Village-level entrepreneurs or FPOs to deliver services to the farmers in a phygital model.