We are entrusted with the task of providing adequate and nutritious food for 9 billion people by the year 2050. Buffeted by the headwinds of global climate change, systemic supply-chain disruptions, geopolitical tensions, soaring food prices, and inflation, agriculture companies and farmers are increasingly turning to technology to provide efficient and sustainable food security for our planet’s population. While technology’s contribution to agriculture has been growing in the past few years, 2023 will see AI particularly emerge as a critical enabler and driver of digital transformation of the sector.

I believe that AI can help bring transformation in agriculture in 2023 and beyond.

Timely and actionable intelligence from satellite data

We are in the midst of a revolution when it comes to satellites and in the modality of the data being captured and recorded by them. Satellites today have become ubiquitous and getting relevant spatial and spectral resolutions is increasingly very affordable; with the data being accessed even by smaller growers. Ushering in unprecedented standards in quality data, their images are at high-resolution and frequency of updates is near real-time. There are recent satellites that were launched for water and topography to manage fresh water for drought applications, for carbon mapping and spectrometers than imaging sensors.

AI models with right knowledge will be the key then to deciphering and making sense of this huge amount of rich data and translating it into actionable intelligence for farmers and agri-businesses. Integrating the diverse data including differentiated spectral bands, parameters like canopy temperature, soil temperatures and information, light and radiation, hyper local weather can provide advisory through various channels for growing crops reliably and more precisely.

Regenerative Agriculture, Sustainability and Biodiversity

As we are catering to producing food for the growing population, it is possible that we add more stress on the environment and natural resources. At the same time, land surface temperature and the ambient air temperature is increasing in certain regions while other regions are facing extreme rainfall. The prior knowledge created can help cater to the changes in crop cycles, seasonal cycles and growing patterns. Adoption of integrated farming practices can help regenerate soil, and reduce carbon and water footprint, and produce locally from farm to fork. With the recent COP15 resolutions passed by countries, the availability of high resolution satellite images can help government and development agencies to map out biodiversity hotspots and forest regions. Crops can be interlaced with existing trees symbiotically to forest ecosystems and food forests. Many countries are consciously making decisions in reducing the deforestation footprint and forest intrusions during procurement.

Integrating knowledge from diverse players in the ecosystem

Currently, the ecosystem comprising farmers, seed manufacturers, farm businesses, equipment manufacturers, input and processing companies, agronomists, governments, developmental agencies, and technology providers all operate in silos. The innovative research and rapid advances can be integrated and cross-leveraged for bringing multiplier impact on the entire ecosystem. The challenge is to bring these different types and layers of data and knowledge together in a contextual way that can provide accurate advisories to farmers. 

For example, it may be possible to estimate the potential risk of disease in a particular region or area in advance, and intervene. However, if rain is expected in the next few days, it would make any effort ineffective. Custom built models can help layer these two pieces of data on top of each other, and provide an advisory to the farmer to postpone any interventions until the weather clears so that the input usage can be optimised.

Climate Smart advisory using weather and smarter lending and insurance

The weather and seasonal changes holds the key for many marginal farmers. Despite recent advances, weather forecasts today, are less accurate for future days and the reliability is for the next few days at best. We expect that weather predictions become more accurate for not only fortnight forecasts but also for seasonal forecasts which will immensely aid in planning and risk mitigation.

Crop insurance and access to credit are critical but often overlooked, constituents of agriculture. We can now infer plot boundaries directly from satellite data throughout the crop growing cycle and we anticipate that these could be used to automate the cadastral map updation and land ownership identification. Thanks to this, governments and developmental organisations can undertake insurance underwriting and risk monitoring. For farmers, customised crop-insurance products can be launched for ensuring climate proofing of production by looking at the long-term climate patterns and current weather.

Pradhan Mantri Fasal Bima Yojana (PMFBY), a flagship Indian government scheme which involved the rollout of crop insurance scheme encompassing 250,000 gram panchayats across India, offers a preview of things to come. The project required the states to assess the Crop Cutting Experiment (CCE) in every Gram Panchayats: Cropin’s AI/ML-powered predictive intelligence platform helps the government to conduct this mammoth exercise which includes regional-level crop identification and phenological growth stages, health, and growth analysis, crop yield forecasts, and real-time insights into on-ground CCE progress. This flagship project helped government agencies to carry out insurance underwriting and risk monitoring, introducing custom-made crop insurance products and schemes and yield index monitoring at a national level.

Region & country level food security planning

By monitoring land usage using our dynamic LULC and by identifying crops, AI can help governments to better plan the estimated crop and food stock supply, and map it against the projected demand. Accurate estimation of the yield level, helps avoid supply-demand imbalances and prevent food crises like the one we are facing globally now. The soaring prices in oil seeds and certain cereal crops has made it very clear that the problem of one country becomes the problem of the world.

Recently Cropin carried out a wheat production analysis for the Nigerian government. The company used satellite imagery data, real-time and historical weather data, farm geo-tags and advanced AI models to estimate the total wheat acreage area, yield per hectare, total yield and crop yield prediction across 15 states of Nigeria. This kind of analysis helps us in identifying regions that are traditionally very agrarian but might not be able to produce due to current geopolitical situations.

The agricultural use cases of AI and ML are massive and we are yet to leverage the full potential of these emerging technologies in this segment. Tapping the vast promise of AI in agriculture will help the world produce food for a growing population sustainably and equitably.

Photo by Aabir Ahammed on Unsplash