Up to 60-70% of input cost goes into managing disease and pesticides, according to experts
AI-based agritech solutions provide intelligence around crops, fertilisers, climate and soil to help farmers reduce wastage and losses
There is a dearth of farming data in the country which hinders research in agritech, said Ninjacart cofounder Thirukumaran Nagarajan
For centuries, Indian farmers have been toiling in the fields, and relying on intuition and traditional wisdom to make the most of a bad situation. Besides tending the soil, crops, and livestock, farmers have to deal with climate change, the shifting market conditions and pray that the harvest is good enough to earn some returns. The uncertainties around climate, water and pests among others has made life a struggle, with farmer suicides rising every year.
But agritech startups are increasingly stepping in to save income, lives and livelihoods in India’s agricultural sector. While supply chain is one of the biggest inefficiencies in farming, intelligence about crops, fertilisers, climate and soil are also key. This not only increases the input cost, but also the chemical residue from fertilisers can affect the crop, and it can even be present in the food that we consume.
Leveraging technology such as artificial intelligence (AI), machine learning (ML), internet of things (IoT), analytics and blockchain becomes crucial for the farming industry in this regard. It not only promotes intelligent agriculture, but also cuts down their stake and input cost by 40-60%, thereby increasing their saving, yield and protecting the fertility of the land.
According to MarketsAndMarkets, the AI in agriculture market is expected to reach 4 Bn by 2026, growing at a compound annual growth rate (CAGR) of 25.5%, from $1 Bn in 2020. The major driving factors of AI in agriculture include growing demand for food, and dwindling natural resources.
The growing use of technology, especially computer vision tech for agriculture applications including plant image recognition and the skyrocketing demand for monitoring and analysing crop health are some of the essential factors contributing to the growth of the AI sector and solutions based on computer-vision technology.
In India, however, AI in agritech is still in the nascent stage. The technology has the potential to radically improve agriculture advisory, support farmers and can bring objectivity and transparency to post-harvest value chains. Making use of the opportunity, many startups are burgeoning in the space.
“We have seen AI adopted across post-harvest quality platforms like Intello Labs, used to guide field robots with Tartansese, embedded in aquaculture IoT systems like Eruvka, and in horticulture IoT systems with predictive algorithms like Fasal,” said Mark Kahn, managing partner at Omnivore, an agri-focused venture capital firm. Besides this, some of the other startups also include CropIn, Niruthi, Aibono and NinjaCart among others.
According to DataLabs, the agritech sector recorded a total funding of $244.59 Mn in 2019, an increase of over 350% in the amount of funding, compared to previous year. Several agritech stakeholders told Inc42 that whole startups have expanded rapidly in the last few years, the sector, however, has been held back due to lack of structured data, fewer large investments, poor infrastructure and policy lag.
How Startups Are Leveraging AI?
Today, agritech startups use a mix of supervised and unsupervised machine learning models, from neural networks to machine learning, IoT-based sensor data mining and pattern recognition algorithms to provide a variety of decision making solutions to farmers, markets, resource managers and policy makers. The models can be used to both describe and diagnose crop type, crop conditions, to make predictions and schedule farming operations, such as irrigation and harvesting.
For instance, machine learning models are used to map various crop areas from satellite images. Algorithms such as random forest and support vector machines are commonly used for automated crop type identification. Maps of crop areas are used for crop area assessment, crop yield forecasting and resource use optimisation.
In precision farming, machine learning can be used to monitor crop stage by identifying flowers and fruits on plants and use the information to schedule fertilisation, pest control and other crop practices. Crop grading also takes advantage of computer vision and deep learning algorithms, which can be trained to identify the size and quality of fruit and other produce and automatically sort it by grade.
In addition to this, AI models can also be used to monitor livestock, identifying sick animals, schedule feeding and, in the case of dairy farms, monitor milk production. Moreover, a different set of natural language processing AI models can be applied to set up chatbots to increase communication along the supply chain and identify any issues. The list is endless.
Speaking to Inc42, Ananda Verma, founder and CEO at Fasal, said that they use solar-powered IoT devices which collect data in real time and transmit it to the cloud. The data collected is then crunched and analysed using an AI platform to predict various stages of the crop from the seed stage and climate change. Once analysed, it provides preventive action to farmers, thereby helping them achieve quick yield and better produce.
Based in Bengaluru, Fasal provides farm-specific, crop-specific and crop stage-specific actionable insights to farmers. The insights could be in terms of irrigating the crop, predicting possible disease, and spraying of pesticides, etc. The company has 25 crops in its portfolio, including grapes, pomegranate, capsicum, tomato and other cash crops. Currently, it is present in Chhattisgarh, Madhya Pradesh, Maharashtra, Karnataka, Telangana and Andhra Pradesh.
“There is a dearth of data in the country which hinders research and therefore causes a problem for the use of AI in agritech,” said Thirukumaran Nagarajan, cofounder and CEO of supply chain major Ninjacart. Further, he said that the influence and role of AI in agriculture is very slow to begin with, and it is too early to comment about whatever minimal inroads it has made in the world of agitech.
“Once we are able to get banks of data, we can easily apply AI which will help us solve many problems related to agriculture,” said Thirukumaran.
Moreover, in most cases, agritech startups are still figuring out a viable real-use case study that can be implemented in the practical world, he added.
Bengaluru-based B2B fresh produce supply chain startup Ninjacart said that it has seen considerable progress on the application of AI in farm management, but they are yet to experience any major impact in the area of large scale growth before going mainstream. “However, we will see real use cases evolve in the coming years,” said Thirukumaran.
How AI Is Transforming Agriculture In India?
“AI is helping farmers to analyse and interpret conditions such as the condition of the weather, water use and soil conditions or temperature information collected from their farm for better decision making,” according to Mallikarjun Kunkunuri, CEO of Niruthi Climate and Ecosystem Services. Further, he believes AI technology can help the farmers to optimise planning and strategizing along with generating better yields by best hybrid choice, crop choices and efficient utilisation of the resources.
Based in Hyderabad, Niruthi focuses on strengthening the resilience of farmers to climate change, and offers AI-backed crop insurance and irrigation solutions. Bringing crop insurance to every farmer in India requires scalable technologies that can improve the efficiency of ground data collection for crop monitoring and yield assessment, as well as provide rapid assessment through the proper use of satellite data at large scales. “By providing accurate and timely data, we hope to reduce uncertainty and improve efficiency in claim settlements,” said Kunkunuri.
Similarly, timely data acquisition at scale and the increase of data mining applications also helps banks and government agencies to plan and deliver on necessary inputs and in creating appropriate credit policies, he added.
In addition to this, the company also helps to identify the use of water in agriculture to maximise farm yields, reduce water use and conserve power. Also, it uses machine learning models to produce hyper-local weather information that is used to provide advisories about impending pest outbreaks, and when to perform preventive pest control.
Krishna Kumar, CEO of CropIn, a full-stack agritech that provides SaaS solutions to agribusinesses globally, says that one of the greatest blessings of artificial intelligence in agriculture is the ability to continually monitor crop health and detect any pests or diseases that could damage the crop.
“AI in farming comes as a godsend to farmers who can now get alerted about the slightest change in their crop with as little as a message on their phones,” added CropIn’s Kumar.
Further, he said that smart agriculture powered by AI combines inputs from satellite imagery, weather and ground data to allow complete digitization of farms at a plot and regional level. “This not only provides complete visibility throughout the cultivation process but allows end-to-end traceability of the produce from farm to fork,” he added.
Ninjacart’s Nagaranjan said that AI in agritech will experience a huge momentum shift when there is rapid AI adaptation. This will eventually lead to farm automation, thereby reducing dependency on labour and bringing down costs. Overall, this will also improve efficiency on multiple fronts, especially lower emissions and wastage. “The major areas we see AI impacting Indian agriculture are in pest control, weather management, crop quality, continuous monitoring and learning,” he added.
The Future of Agriculture And AI
According to agritech experts the future trends for agritech startups in India include increased use of AI in support to crop insurance and lending, where AI will be increasingly used for creating a network of on the ground monitoring to improve training and validation of scaling solutions that use satellite data.
Given the large heterogeneity in crop grown, varieties, staggered sowing times, and the fact that traditional crop cutting experiments for yield assessments are time consuming, to bring insurance and finance to every farmer there will be a greater scope to apply AI for rapid ground yield assessment.
With increased use of IoT sensors for continued field condition monitoring, especially for cash crops, there’s a large need for yields to reach full potential across the farms. Another trend in agritech startups is increase in storage facility and supply chain traceability to ensure quality of the crop distribution, access to market at a fair price, reduce food wastage and buffer against climate fluctuations. Use of AI in this space can also improve matching of supply and demand and provide real time visibility into the inventory.
“AI comes as a great boon to the agricultural sector which is heavily dependent on climatic conditions which are often unpredictable. More and more use cases of AI in agriculture are likely to show up in the near future because of the immense value it can add,” said CropIn’s Kumar.
Further, he said that new techie farmers are ambitious and are moving towards indoor farming. It is a technique of growing crops or plants, typically on a large scale, entirely in a packed environment. “This way of farming often implements growing methods like hydroponics and leverages artificial lights to provide plants with the nutrients and light levels required for growth. AI-powered indoor agriculture is tempting a whole new breed of farmers now,” Kumar said.
Last year, 80 Acres Farms, one of the pioneers in indoor growing opened the world’s first fully-automated indoor growing facility last year. The company’s AI-driven technologies monitor every step of the growing process.
“It’s too early to predict the impact of AI in agitech. However, we will definitely see agritech startups growing exponentially with AI playing a major role across the spectrum in the near future,” concluded Ninjacart’s Thirukumaran.
