How data and artificial intelligence can drive Asia’s sustainable farming future

I grew up in a little hill station town called Shillong. It’s a beautiful area in India’s northeast corner that’s called the “Scotland of the East” and not far from where they grow tea in Assam. It’s about 50 miles from the rainiest place in the world near the border with Bangladesh.

The economic inequality I saw growing up taught me to try to understand what the wider community is dealing with; to understand what so many people are doing every day just to survive. We can all do more and give back.

With only one-fifth of the world’s agricultural land, Asia hosts more than half of the global population. Climate change and increasing food prices are critical threats to long-term food security – more than 1 billion people lack access to sufficient food across the region. The situation is part of a global trend that the United Nations calls an “unprecedented food emergency”.

Meeting Asia’s food demand will be challenging due to slowing gains in agricultural productivity, overexploitation of natural resources and increasing water scarcity. As the region becomes more urban and prosperous, food prices will continue to increase unless supply can keep up with demand. To feed Asia’s growing population more sustainably and efficiently, the way food is produced must change.

Technologies like artificial intelligence (AI), sensors and drones can help increase agricultural productivity, food safety and agri-food system sustainability. From the soil – where better farming practices can mitigate climate change – to the shelf – where customers look for products with minimal carbon footprint – Asia’s agriculture and food value chain is primed for innovation.

Microsoft’s goal is to democratize data-driven insights to help all farmers and organizations achieve more and transform the agri-food value chain to become more productive and more transparent, and drive shared value all the way back to producers.

Microsoft recently announced Microsoft Azure Data Manager for Agriculture in preview. What began with Project FarmBeats, an ambitious research initiative to collect and transform agricultural data, has now evolved into a timely commercial solution. Azure Data Manager for Agriculture extends the Microsoft Intelligent Data Platform with industry-specific data connectors and capabilities to connect farm data from disparate sources.

For example, industry leader Bayer’s FieldView platform ingests data from Azure Data Manager for Agriculture’s satellite and weather pipelines to produce insights on potential yield-limiting factors in growers’ fields. Bayer has also developed a set of digital solutions available to enterprise customers that provide timely insights on crop health, weather forecast, crop growth tracking, and more.

Data-driven agriculture is also a foundational component of Land O’Lakes’s digital offerings, including the Truterra sustainability tool. This innovative digital service provides farmers with insight into how different agricultural practices impact water, nitrogen and carbon on a farm, enabling them to track their soil’s carbon sequestration, among other applications.

Meanwhile, BharatAgri, an Indian agricultural start-up, leverages data from satellite imagery to monitor crop health and analyze farms as small as 1/40 of an acre. This year alone, more than 50,000 farmers are expected to receive satellite images of their farms from BharatAgri, which will help reduce crop losses on over 100,000 acres of farmland.

Globally, data-driven agriculture has been gaining momentum as one of the most promising approaches to addressing the food security challenge. According to the International Food Policy Research Institute, data-driven agriculture techniques can increase farm productivity by as much as 67% by 2050, while simultaneously cutting down on agricultural and food losses.

However, the high costs of adopting new technologies can also be a barrier for low-to-middle-income countries. This is especially critical for Asia where smallholder farmers are the major group, with 450 million producing more than 80% of the food consumed in the region.

Data-driven agriculture starts with collecting information about the farm, which can be especially difficult in rural communities that lack digital infrastructure. This data is obtained from a variety of sources, including sensors, drones, tractors, weather stations and satellite imagery, making affordable internet connectivity a necessity.

Over time, all these data streams can be used to indicate useful practices and make suggestions based on previous crop cycles, resulting in higher yields, lower inputs and fewer environmental effects.

For maximum impact, the right data needs to be leveraged for the right purpose and at the right time. But the enormous size and complexity of agri-food systems, coupled with their fragmented nature, pose challenges to unlocking big data’s potential economic value, which is projected at over $100 billion in Southeast Asia alone. To secure more inclusive growth, smallholder farmers must be empowered to participate in modern agri-food value chains.

With so much agriculture-relevant data generated across the farm – as collected using sensors in the soil to satellites orbiting the earth – many farmers and organizations just don’t have the right resources to harness these large amounts of data effectively. The agri-food value chain also comprises a complex system of stakeholders and activities, so datasets remain siloed without interoperable systems for managing them.

The good news is AI can help in breaking down data silos and transform a huge amount of complex agricultural data into actionable insights. Based on years of study in Microsoft Research and thought leadership established with Project FarmBeats, we released FarmVibes.AI – a suite of tools aimed at guiding decisions at every phase of farming.

FarmVibes.AI workflows, which are run on cloud technology, provide rich predictive and prescriptive insights on soil health, weather patterns, carbon sequestration, waste tracking and more. They can help farmers predict the ideal amounts of fertilizer to use and where to use them, forecast temperature and wind speeds, and inform when and where to plant and spray, among a range of other agricultural applications.

These insights won’t be possible without data fusion techniques powered by AI, which combine data streams from sources like weather station data, drones or satellite imagery. This also helps to unlock cost-efficient gains and improve accessibility to digital agriculture solutions. Leveraging AI can reduce the number of sensors and drones needed, thereby reducing the cost of on-farm hardware.

The natural language capabilities of today’s large language models can also help make these technologies more accessible for farmers who are not as technologically savvy. Through Project FarmVibes.Bot, for example, smallholder farmers can communicate simply and effectively to query data or relay insights.

The availability of affordable internet-connected sensors – underpinned by cloud technology and AI – can empower farmers to capture and track operational data, whether it’s from the soil, equipment or livestock. Using AI technology, farmers can generate insights based on that data to apply precision agriculture or predictions for improving yields, while conserving precious resources.

Data and artificial intelligence can also augment farmers’ special knowledge and intuition of their farms to make much more informed decisions. With better insights, farmers can increase harvest and production efficiency, reduce food waste, create nutrient dense and high-quality products, reduce environmental impact, and provide transparency to stakeholders.