The agriculture industry has always been at the forefront of innovation and technology, and in recent years, the integration of artificial intelligence (AI) has become a hot topic among industry experts. With promises of increased efficiency, yield, and profitability, many are looking to AI as a source of productivity greater than what was experienced during the “Green Revolution.”
In a recent interview with MIT’s Sam Ransbotham, Land O’ Lakes’ Teddy Bekele made these startling comments:
Sam Ransbotham: So, with all these data, artificial intelligence, and machine learning models you’re using, somehow, you’ve gone in this fourth revolution or fourth … I can’t remember what you called it, fourth wave?
Teddy Bekele: Revolution. Yes.
Sam Ransbotham: What’s next?
Teddy Bekele: With some of the capabilities, both the biotechnology as well as the software technology component of it, there’s farmers that can get up to 540 bushels per acre. However, as this article explores, the reality is more complex, and there are no easy solutions when it comes to AI in agriculture. Despite this, the potential benefits of AI cannot be ignored, and it remains a crucial tool in the ongoing quest to feed an ever-growing global population.
Every October, a global cadre of business and political leaders descend on Des Moines, Iowa. But the likes of Tony Blair, Kofi Annan, and Bill Gates do not come for the presidential caucuses that make the state famous. They are drawn to the sleepy state capital for the annual World Food Prize, the top award celebrating innovators in the agriculture sector.
The World Food Prize has its roots in the Third Wave or Green Revolution during the “1960-1980s” a period of rapid advances in agricultural productivity. Two agronomists – Norman Borlaug in America and Yuan Longping in China – are largely credited with spurring the Green Revolution by pioneering new varieties of wheat and rice. But new plant varieties were also buttressed by other technological advances that allowed farmers to produce more food on less land.
There is no telling how many millions of people were saved from starvation due to these advances. But today, the system pioneered by Borlaug, Longping, and other innovators appears to be bumping up against its production limits. Also dogging the industry, the sector accounts for about a third of global greenhouse gas emissions. Fertilizer and pesticide runoff, land use changes, and heavy draws on freshwater supplies are also increasing the industry’s strain on the environment. Weather volatility has also put a constraint on the industry as key growing areas, like California, veer wildly from drought to floods.
Steady increases in demand, driven by a burgeoning world population, will continue to provide ample market opportunities across the food and agriculture value chain. Yet, climate change and resource scarcity will negatively impact the ability of farmers to deliver consistent yields. If United Nations population estimates are correct, there will be nearly 10 billion mouths to feed by 2050.
Artificial intelligence (AI) is therefore one critical tool that will help the food and agriculture sector to effectively meet global demand. Also, AI and agricultural technology (AgTech) more broadly, could help turn more agricultural production from a climate liability into a climate asset. AI will likely make key contributions in optimizing the use of resources, addressing agricultural labor challenges, and discovering new agricultural technology. That is one reason why the AgTech sector attracted nearly $30 billion in investment last year.
Surveying AI and Ag landscape
As the agriculture industry continues to explore the potential benefits of AI, it is important to take a step back and get a bigger picture of what’s happening. While the use of AI in agriculture is still in its infancy, with many challenges and barriers to widespread adoption, there are also several promising developments and success stories that point to the potential of AI to revolutionize the industry.
Narrow forms of artificial intelligence, powered by advances in machine learning algorithms, are upending traditional business models throughout the economy. AgTech founders are at the forefront of using narrow AI models to enhance agricultural production. The capabilities that make AI such a revolutionary tool – learning, adapting, and generating unique analytics and insights – are being put to powerful use in the agricultural context.
The opportunities are vast across each industry segment, from plant and animal agriculture to marketing and distribution. Realistically, there are several challenges to the increased adoption of AI tools in agriculture.
One of the biggest hurdles facing the integration of AI in agriculture is data collection and analysis. Agriculture generates vast amounts of data, from soil sensors to satellite imagery, but the industry has traditionally been slow to adopt digital technologies. This has resulted in a lack of standardization and interoperability among data sources, making it difficult to integrate and analyze data in a meaningful way. The good news is that there are several initiatives underway to address this challenge, such as the Global Open Data for Agriculture and Nutrition (GODAN) initiative, which aims to promote open data policies and interoperability among data sources.
Another challenge is the complexity of agriculture systems, which can be difficult to model and predict using traditional AI algorithms. Location, soil, plant, and weather are just a few of the variables confronting AI adoption. Recent advances in machine learning and deep learning are allowing for more accurate and sophisticated models that can consider the complexity of agriculture systems. For example, researchers are using machine learning algorithms to analyze crop yields and predict the impact of climate change on agriculture.
Other challenges include the disaggregated structure of the farming sector, remote locations of many farms, and – in developed economies – an older-than-average workforce.
If data, connectivity, and computing power are the building blocks of AI, development in agriculture still has a long way to go. But the sustained investment in the space demonstrates the eye-popping potential that venture capital investors are seeing. As an example, the dramatic rise in the value of an acre of Iowa farmland to $30,000 demonstrates the appetite.
Despite these challenges, there are several success stories that demonstrate the potential of AI in agriculture. For example, precision agriculture technologies that use AI to analyze data from sensors and drones have been shown to increase crop yields while reducing inputs such as water and fertilizer. AI-powered autonomous farming machines are also being developed that can perform tasks such as planting and harvesting crops with greater efficiency and accuracy
Rob Leclerc, the founder of AgFunder, a specialized media outlet focusing on AgTech, sums up the outlook nicely: “As the enabling technologies mature, middleware providers will develop ways to integrate all of these advances, and AgTech companies won’t have to boil the ocean every time, in order to build an app. They can basically identify the data or communications and machine intelligence capabilities they need, buy them, plug them in, and spend time on solving a problem with their product.”
Any way you slice it, agriculture presents a fascinating canvas for tracing the likely development and implementation of AI business models. We now turn to the venture capital investment data for a closer look at three of the biggest potential uses for AI in agriculture.
Optimizing scarce resources
One of the key ways that AI is being used in agriculture is to optimize the use of resources such as water, fertilizer, and pesticides. Agriculture is a resource-intensive industry, and reducing waste and increasing efficiency are crucial for both economic and environmental reasons. AI-powered precision agriculture technologies are allowing farmers to target resources more accurately where they are needed most, reducing waste, and increasing yields.
For example, AI-powered irrigation systems can analyze data from soil sensors, weather forecasts, and crop models, to determine the precise amount of water needed for each crop and adjust irrigation accordingly. This not only reduces water waste but also ensures that crops receive the optimal amount of water for growth and yields.
Similarly, AI-powered nutrient management systems can analyze soil samples and crop data to determine the precise amount of fertilizer needed for each crop and adjust application rates accordingly. This not only reduces fertilizer waste but also ensures that crops receive the optimal nutrients for growth and yields.
AI can also be used to optimize pest management, reducing the need for harmful pesticides. For example, AI-powered pest detection systems can analyze crop images and identify pests or diseases at an early stage, allowing farmers to take targeted action to prevent further spread without having to apply pesticides across entire fields.
Overall, the use of AI to optimize the use of resources in agriculture has the potential to reduce waste, increase efficiency, and improve yields while also reducing the environmental impact of agriculture. As technology continues to advance, we can expect to see even greater benefits in the years ahead.
Using AI to improve the efficiency and productivity of farming represents the lowest-hanging fruit for AI’s contribution to the agriculture sector. No matter the product, farmers share twin goals: minimize inputs and maximize yield. AI models are being used to supplement farmers’ knowledge and experience as farming becomes more complex. Water scarcity, high fertilizer and input prices, and volatile commodity markets combine with extreme weather to make farming a tough prospect. Considering that most farms operate like small businesses with few employees, AI-powered digital assistance can help reduce costs and increase profitability.
AI models can help farmers and agricultural producers manage the complexity in a few distinct ways.
First, AI is enabling earlier identification of potential issues with plants and animals. As technology develops, AI systems will be far better at identifying important characteristics than humans. Companies are springing up to capitalize on the potential. Cromai in sugarcane, Agrisolus in poultry, and XpertSea and Aquabyte in aquaculture are all applying proprietary AI models to better predict plant and animal health. Like a doctor to a patient, a farmer or grower needs the best available information to act decisively. Early detection and intervention are key.
Another key way that AI is being used in agriculture is to optimize the use of resources such as water, fertilizer, and pesticides. Agriculture is a resource-intensive industry, and reducing waste and increasing efficiency are crucial for both economic and environmental reasons. AI-powered precision agriculture technologies are allowing farmers to more accurately target resources where they are needed most, reducing waste and increasing yields. The same models that identify potential issues can also provide targeted, field-specific recommendations. The models mimic all the complexity found in the real world, such as soil type, crop type, and weather patterns.
For example, AI-powered irrigation systems can analyze data from soil sensors, weather forecasts, and crop models to determine the precise amount of water needed for each crop and adjust irrigation accordingly. This not only reduces water waste but also ensures that crops receive the optimal amount of water for growth and yields. Similarly, AI-powered nutrient management systems can analyze soil samples and crop data to determine the precise amount of fertilizer needed for each crop and adjust application rates accordingly. This not only reduces fertilizer waste but also ensures that crops receive the optimal nutrients for growth and yields.
Crucially, these technologies may not need to scale on their own to be successful. Established agribusiness equipment makers and input suppliers have shown a willingness to splurge on AI technology that can be integrated into their existing businesses. Tractor maker John Deere spent over $300 million to acquire Blue River Technology, which developed an AI model that only applies pesticides to problematic weeds (as opposed to the current industry practice of spraying broadly across an entire field). The innovation promises to save farmers money and improve environmental outcomes.
Here are a few other VC Ag AI companies in the crop and soil management space:
Taranis secured $40 million in Series D funding. Taranis uses AI-powered high-resolution imaging and machine learning to detect early signs of crop diseases, pests, and nutrient deficiencies, enabling farmers to address issues before they become detrimental.
Trace Genomics recently received a $17 million Series B funding, to advance their AI-driven soil analysis platform that helps farmers understand soil health, fertility, and microbial communities. This allows farmers to optimize fertilizer usage and improve crop Yield.
Stenon, has a real-time soil-sensing solution without the need for a lab, and has raised a $20 million Series A funding round.
Like all narrow AI models, robust data sources are key. AI models are being enabled by a new wave of satellite imagery and remote sensors. These technologies have vastly expanded the amount of on-farm data generation, which can be fed into AI models to create virtuous learning cycles.
Mitigating agricultural labor challenges
Agriculture is a labor-intensive industry, and labor shortages and rising labor costs are major challenges facing farmers worldwide. However, AI has the potential to mitigate these challenges by automating tasks that were previously performed by human workers.
COVID-19 laid bare just how dependent agriculture in developed economies has become on migrant labor. Border restrictions stemming from the pandemic largely halted seasonal migrant farm workers, imperiling harvests in Europe and North America. To make matters more challenging, anti-immigration sentiment continues to flare, despite the substantial contributions migrant labor makes to the agricultural economy.
Given the circumstances, agribusiness leaders are understandably keen on finding alternatives that can help overcome acute labor shortages. Investors are betting that AI-powered robotics can provide at least part of the solution through the development of autonomous farming machines. These machines can perform tasks such as planting, weeding, and harvesting crops with greater efficiency and accuracy than human workers. This not only reduces labor costs but also reduces the physical strain on workers and can improve the safety of agricultural work.
Two California-based firms, Verdant Robotics and Farmwise, hauled in $46.5 million and $45 million respectively to help accelerate their development of robots that can harvest crops and complete other on-farm tasks.
Although, not all crops are well-suited for current forms of mechanical harvesting. Fragile fruits and vegetables, such as berries and tomatoes, still require a human touch. Someday soon, one of these upstart AI firms may help cross that barrier (after all, isn’t that what AI is all about?). In the meantime, robots programmed with machine learning algorithms will prioritize the harvesting of more durable crops like tree nuts. AI can also be used to optimize workforce management by analyzing data on worker productivity and identifying areas for improvement. For example, AI-powered systems can monitor worker movements and identify patterns that may indicate inefficient work practices. This can help farmers to optimize their workforce and improve productivity.
The dynamic between AI and labor in the agriculture sector does underscore a critical point about AI technology: Narrow AI is likely to augment, rather than replace, human labor. Ideas of machines taking jobs are often in vogue, but the general AI systems that can outperform humans on an array of cognitive and abstract reasoning tasks are a long way off. The AI robots zipping around farmers’ fields will still need humans to input important parameters, check work, and troubleshoot any issues. In the same way that automated teller machines (ATMs) did not completely replace bank tellers, AI models will not eliminate the need for farm workers. However, they are likely to free up employees for more value-added tasks and reduce the grueling, manual labor currently required for harvesting some crops.
New production methods & supply chains
The final critical area venture capital investors are targeting for AI and agriculture is the discovery and development of new technologies. With the vast amounts of data generated by modern agriculture, AI can help to identify patterns and insights that might otherwise go unnoticed. This can lead to new agricultural approaches with transformative potential. In this category, AI algorithms serve as high-powered research assistants. If Google has a high school diploma, think of these search models as having a Ph.D.
Take EarthOptics for example. The firm received a $10 million series A investment to help farmers generate a new income stream from carbon sequestration. Some types of healthy soils store carbon in the ground and can serve as a carbon sink much like forests. Governments and academia have taken note. Practices that promote “carbon farming” are proliferating across the agriculture industry, and many farmers hope they can get paid by large emitters for storing more carbon in their soils. The major challenge is accurately measuring and assessing how much carbon is present in the soil and how long it stays there. That is where EarthOptics comes in. The company is developing an AI model that uses satellite imagery and remote sensing to measure carbon and provide farmers with a clear picture of their sequestration potential. Without advances in AI, this level of unique insight into a farmer’s specific field and soil type would not be possible.
Another tranche of investment is flowing into firms that are on the cutting edge of developing the so-called bioeconomy. The goal of these firms is to identify and commercialize genetic traits in plants and animals that have positive economic and environmental benefits. Farmers and agribusinesses see the potential for these traits to improve nutritional content in food, make plants more resilient to drought, and reduce the greenhouse gas emissions of livestock.
Seed-X, Traitology, and Bright Seed combined for over $100 million in recent funding rounds for this exact type of AI-powered research. For example, Bright Seed’s Forager® platform can sift through thousands of traits to identify bioactive compounds with potential human health benefits. In all, the three funding rounds were a small fraction of the larger investments made into the agricultural biotechnology and biomaterials segment. According to AgFunder, that number totaled nearly $5 billion in 2022.
Lastly, AI is playing a crucial role in optimizing agricultural supply chains, ensuring that perishable goods reach their destinations efficiently and with minimal spoilage. Venture capital investments have fueled the growth of several startups in this space:
Full Harvest raised $27 million in Series B funding to enhance its AI-powered B2B platform that connects farmers with food and beverage companies to sell imperfect or surplus produce, reducing waste and improving supply chain efficiency.
Strella Biotechnology has developed an AI-driven IoT sensor system to monitor fruit ripeness and quality in storage facilities, enabling better inventory management and reducing spoilage. They have also recently closed a $8 million Series A round.
AgShift‘s AI-based quality assessment platform streamlines the inspection process for fresh produce and seafood, allowing buyers and sellers to agree on quality and pricing more efficiently. The startup secured $5 million seed funding to expand its technology.
The integration of AI in agriculture is a complex and multifaceted topic that requires careful consideration of both the potential benefits and challenges. While AI has the potential to revolutionize the agriculture industry by increasing efficiency, sustainability, and profitability, there are also significant barriers to widespread adoption, such as data collection and analysis, interoperability, and model complexity.
Despite these challenges, there are several promising developments in the field of AI in agriculture, such as precision agriculture technologies, autonomous farming machines, and plant breeding optimization. These developments highlight the potential of AI to discover and enable new approaches and production methods in agriculture.
To fully realize the potential of AI in agriculture, it is important for stakeholders to work together to address the challenges and promote the adoption of digital technologies. Initiatives such as GODAN, which promote open data policies and interoperability, are a step in the right direction. It is also important for farmers and researchers to collaborate on the research and development of AI applications that meet the specific needs of agriculture.
AI is a powerful tool that can lead to increased efficiency, sustainability, and profitability when applied in a thoughtful and strategic manner.
While the image of the overall-wearing, pitchfork-wielding farmer may still be prevalent in the eyes of the public, the reality is that the people who grow, harvest, and retail our food are likely to be at the forefront of utilizing the most advanced technology available in the world.
Thomas Jefferson once called agriculture our “wisest pursuit.” He could not have envisioned just how wise the sector may become, with the help of AI.