Agricultural Artificial Intelligence: Because it’s a Hungry World Out There

It is expected that by 2050, the Earth’s population will grow to 10 billion people or more. While it is possible that some limited new land can be brought under cultivation, it is clear that the world’s farmers will have to produce far more food with only scant increases of tillable land. Artificial Intelligence (AI) will be the farmer’s tool in achieving this goal.  

As described in an article published by the University of Florida^[1], farmers can now collect and analyze vast amounts of data about their crops. This includes weather patterns, soil health, and plant growth. Through the use of AI, modern farmers can improve their methods, making it possible to improve their crop yields, reduce waste and ultimately increase profitability.  

The Pivotal Role of AI in Modern Agriculture

As we face the daunting challenge of feeding a burgeoning global population, the role of AI in agriculture becomes increasingly pivotal. According to experts from the University of Florida:

"With the help of AI to optimize farming practices, farmers can improve their crop yields, reduce waste, and increase profitability. Accurate and timely data can make the difference between a successful or unsuccessful crop cycle, making it crucial for farmers to leverage technology and data to their advantage." 

This insight highlights the critical role of AI in enhancing agricultural productivity and sustainability, essential for meeting the food demands of a growing population with limited arable land.

A critical challenge faced by farmers is sustainability. Forbes^[2] reports that fully “eleven percent of global emissions come from agriculture and that almost 40% of food produced in the US ends up being wasted. We also learn that just shy of two-thirds of the antibiotics used today are used to treat livestock that are ultimately intended to feed people and not to treat people and their illnesses directly. This is a direct cause of antibiotic resistance that is now emerging as a major factor in healthcare. Agricultural AI can go a long way towards alleviating these issues and more.  

Agricultural artificial intelligence is a fast-evolving field, with significant new announcements out almost daily. Let’s look at some of the AI generalities that also apply specifically to agricultural AI and also at some of what’s even now available for today’s farmers.  

Machine Learning  

Machine learning is an essential part of AI, including agricultural AI. An article by MIT^[3] quotes AI pioneer Arthur Samuel, defining machine learning as “the field of study that gives computers the ability to learn without explicitly being programmed.”   

Traditional computer programming requires creating complete instructions for a computer to follow based on clearly defined inputs in order to accomplish a task. Machine learning, on the other hand, involves allowing computers to program themselves.  

The first step is to garner a vast amount of data relevant to the need. In general, the inputs might be repair statistics, sensor data, or repair reports. The next step is to study the observed cause-and-effect relationships between the data elements in a dizzying array of combinations. Machine learning then puts the pieces together, and based on the patterns observed, it can predict future relationships with a high degree of reliability.  

One of the things machine learning can do is to distinguish a tiny weed from a tiny plant.  

Destroying Weeds with AI Controlled Lasers  

The LASERWEEDER scours a farmer’s land and employs AI to determine what each entity it encounters is – a nascent crop plant or a weed. The machine, described as a mobile data center, can determine the difference between 40 crops and 80 types of weeds. If a weed is detected, a “killer” laser destroys the weed in milliseconds.   

As described in a video from Carbon Robotics ^[4], this AI-based device can take the drudgery out of farm work. NBC describes it as “a killer robot with an AI brain”, but its victims aren’t enemy combatants; rather, its targets are  – weeds!    

LASERWEEDER video 2 min 3 seconds  

The device, costing $1.2 million, can work 24 hours a day and can replace 30 farm workers who are becoming increasingly difficult to find and hire. In some ways, it can outperform human workers in that it finds immature “weedlings” that are too small for the workers to deal with.  

Because the LASERWEEDER destroys weeds with lasers and not with pesticides, it’s a natural option for organic farmers. It’s also reported that the data gleaned from the weed-killing process will be invaluable to farmers, providing them with real-world insights that will allow them to produce more food economically.  

Technical Details

The LASERWEEDER, a brainchild of Carbon Robotics, is redefining weed control in agriculture. Here's the tech magic behind it:

1. High-Tech Vision and Precision: The LaserWeeder uses 42 high-resolution cameras combined with state-of-the-art computing to distinguish between crops and weeds in real-time. This isn't just a camera snapping pictures; it's a sophisticated AI system with deep-learning-based computer vision models that can tell a weed from a crop with sub-millimeter accuracy.

2. Laser Power: Armed with 30x 150W CO2 lasers, the LaserWeeder is ready to fire every 50 milliseconds. Imagine the precision and speed – it's like having a sniper in the field, targeting only the bad guys (weeds, that is) and leaving the good guys (crops) untouched.

3. Environmental Benefits: This isn't just about zapping weeds. The LaserWeeder's method is a leap towards sustainable farming. By using lasers instead of chemicals, it leaves the soil microbiology undisturbed, unlike traditional tillage. This means healthier soil, healthier crops, and a happier environment. Plus, it's a boon for organic farming, offering an economical path to weed control without herbicides.

4. Efficiency and Cost-Effectiveness: Think about the labor and cost savings. The LaserWeeder can kill up to 200,000 weeds per hour and cover 2 acres per hour at 1mph. It works day or night, in all conditions, significantly cutting down the manual labor and the variable costs associated with traditional weed control methods.

In a nutshell, the LASERWEEDER is more than just a tool; it's a revolution in agricultural practices, aligning with the goals of increased efficiency, sustainability, and environmental responsibility.

Tackling Labor Challenges with AI and Automation

The agricultural sector is facing a significant labor challenge, especially with the growing scarcity of willing workers. This is where AI and automation step in as game-changers.

• Reducing Dependence on Manual Labor: AI-driven technologies, like the LASERWEEDER, are prime examples of how automation can significantly reduce the need for manual labor in farming. These technologies are capable of performing tasks that traditionally required a large workforce, such as weeding, harvesting, and monitoring crop health.

• Enhancing Efficiency and Productivity: With AI, tasks are not only done faster but also with greater precision. For instance, AI-powered drones can monitor crop health over large areas within a fraction of the time it would take humans. This efficiency translates into higher productivity and, ultimately, better yields.

• Addressing Labor Shortages: In regions where there is a shortage of agricultural workers, AI and automation provide a viable solution. Automated machinery and AI systems can operate around the clock, compensating for the lack of human labor and ensuring that agricultural operations do not suffer.

• Shifting the Workforce Dynamics: As AI takes over more labor-intensive tasks, the role of the agricultural workforce is evolving. There's a growing need for skilled personnel to manage, maintain, and supervise these AI systems. This shift is creating new job opportunities that focus more on technology management rather than manual labor.

In essence, AI and automation are not just addressing the labor challenges in agriculture; they are transforming the very nature of farming work. By reducing the reliance on manual labor, these technologies are paving the way for a more efficient, sustainable, and productive agricultural sector.

Generative AI in Agriculture  

As described by NVIDIA^[5], “Generative AI enables users to quickly generate new content based on a variety of inputs. Inputs and outputs to these models can include text, images, sounds, animation, 3D models, or other types of data.” It works by identifying “patterns and structures within existing data to generate new and original content.”  

The most famous example of Generative AI is, of course, ChatGPT.  

As a start, farmers can use CHATgpt directly to get information. There are also specific platforms for agriculture. The Farmers Business Network^[6] (FBM) describes a platform called, simply, NORM. The model concentrates on information of use to farmers. As described, examples of questions that can be asked are:  

• Hey Norm! Why doesn’t gypsum raise soil pH?   
• Would this weekend be a good time to plant soybeans in Crawford County, IA?   
• Hey Norm! Does FBN sell glyphosate?   

As detailed by Ambrook^[7], NORM “is built on OpenAI’s GPT-3.5 model, and uses public data like weather reports, soil data, and product labels to answer ag-related questions. It also taps into FBN’s exclusive agronomic data and assets from the USDA’s National Agricultural Statistics Service.”  

Another example is farmer.Chat^[8]. This video illustrates a farmer employing this Farmer.CHAT to get information about a problem area in his field.   

Farmer.CHAT. by Gooey AI

The rest of the video illustrates the interactions between the farmer and farmer.CHAT, as the latter proposes solutions to the insect problem.  

The potential of generative AI in agriculture is vast and varied. From tackling diseases to optimizing yields and revolutionizing crop breeding, AI is not just a tool but a transformative force in the agricultural sector. As technology evolves, we can expect even more innovative applications that will continue to reshape the future of farming.

Real-World Applications of AI in Sustainable Agriculture

Now, let's dive into how AI is not just a buzzword but a real game-changer on the farm. The folks at the University of Florida are onto something big here. They're talking about AI applications that are not just smart, but also kind to our planet:

1. Smart Watering Systems: Imagine a system that knows exactly when your crops are thirsty. That's what AI-driven irrigation is all about. It's like having a weatherman and a soil expert right in your field, ensuring every drop of water is used where it's needed most.

2. Predicting the Future of Crop Prices: Here's where AI flexes its economic muscles. By crunching numbers on climate trends and market shifts, AI helps farmers make savvy decisions. Less waste, more profit – that's the kind of math farmers love.

3. Self-Driving Tractors? Yes, Please!: Deep learning is bringing the future to farms with autonomous tractors. These aren't your granddad's tractors – they're high-tech beasts that know their way around a field, dodging obstacles and getting the job done with no coffee breaks.

4. Weed Zapping with Precision: Thanks to AI, we're seeing a revolution in weed control. Systems like John Deere's Blue River See & Spray™ are like snipers, taking out weeds with pinpoint accuracy. This means less herbicide on our food and in our environment. It's a win-win for farmers and Mother Nature.

5. Crop Disease Prediction: One of the most promising applications of generative AI is in the early detection and prediction of crop diseases. By analyzing vast datasets, including images of crop fields, weather patterns, and historical disease outbreaks, AI models can predict potential disease outbreaks before they become visible. This early warning system allows farmers to take preemptive actions, reducing the spread of disease and minimizing crop damage.

6. Yield Optimization: Generative AI is also playing a pivotal role in yield optimization. It can analyze data from various sources – soil quality, weather conditions, crop health – to generate recommendations for optimal planting, irrigation, and harvesting times. This not only maximizes yield but also ensures efficient use of resources.

7. Customized Crop Cultivation Plans: Another exciting application is the creation of customized crop cultivation plans. Generative AI can process data specific to a farmer's land, such as soil type, microclimate, and previous crop cycles, to generate tailored farming strategies. This personalized approach can significantly boost productivity and sustainability.

8. Enhancing Genetic Crop Improvement: In the field of genetic crop improvement, generative AI stands as a transformative force. It has the capability to simulate countless genetic combinations, enabling it to forecast traits that enhance crop resilience, nutritional value, and yield. This advanced approach significantly speeds up the breeding process, facilitating the creation of superior crop varieties more efficiently than conventional methods.

So, there you have it. AI in agriculture is more than just a fancy term – it's making farming smarter, more sustainable, and, dare I say, cooler. It's not just about growing more food; it's about growing food the right way.

How Ya Gonna Keep ‘Em Down on the Farm?  

As described in an article posted by the University of Arkansas in The Arkansas Journal of Social Change and Public Service^[9], 73% of the crop farmworker population in the United States are immigrant workers, and about 48% of hired crop farm workers have no work authorization.   

The lack of willing workers, foreign, let alone US citizens, is a tremendous problem facing American agriculture. The answer is automation, and the 185-year-old John Deere company is jumping full steam into the AI revolution.  

Simplifying agricultural automation is the fact that, like the factory floor, a farm is a work environment where everyone knows their job and both act and react in proscribed ways. This is entirely unlike the bedlam of busy city streets. So, while self-driving cars are going nowhere fast, the farm is a far more advantageous environment for driverless vehicles.    

John Deere’s autonomous 8R Farm Tractor.

As described in a report published by CNBC^[10], John Deere has taken up the challenge presented by this opportunity in the form of its 8R Farm Tractor, which doesn’t need a driver and instead relies on AI. Deere has curated hundreds of thousands of images from different farm locations and under various weather and lighting conditions so that with machine learning, the tractor can understand what it’s seeing and react accordingly.   

Challenges and Opportunities  

No matter if you’re a general or a farmer, you don’t want to carpet bomb your opponent. If you’re a general, you want to eliminate enemy soldiers, not civilians. If you’re a farmer, you want to destroy weeds, not food crops.  

Agricultural AI will make it easier to efficiently eliminate weeds, in some cases without any pesticides at all. In other cases, when pesticides can’t be completely avoided, agricultural AI will make it possible to only hit the weeds and avoid the food crops.  

This is important for two obvious reasons. The first is that pesticides cost money. The second is that the fewer pesticides that are used in the vicinity of food crops, the more money the farmer can get for his produce. And, of course, most agree that pesticide-free food is just plain healthier.  

As the cost of AI comes down, and as more and more farmers worldwide have access to it, the world will enjoy better, cheaper and healthier foods. And let's not forget that by enabling more automated farming processes, Agricultural AI will mean that fewer farm workers will be stoop laborers, while more and more of them will emerge as machine operators, repair people and network technicians.  

Wrapping Up  

Perhaps the greatest modern breakthrough in agriculture was brought on by the tragic German scientist Fritz Haber, who pioneered a process to mass produce ammonia, which in turn can be used to produce artificial fertilizer. Before this development, the population of the world was less than two billion, and now it is four times that much, despite the ghastly toll of 20th century wars.  

As much as we talk about the new industrial revolution, we may now acknowledge a new revolution in farming, brought on by artificial intelligence, but unlike the Haber Process, which uses vast amounts of energy and is extremely environmentally unfriendly, the AI revolution is essentially a clean revolution that will allow farmers to do more with less.  

It will either reduce pesticide use or eliminate it entirely. It will advise farmers as to their best pathways into both increased profits and increased food yield. And it will allow the world’s rapidly declining cadre of farmers to feed a hungry world in a more sustainable manner.  

Key Takeaways  

• Agricultural AI will reduce the use of pesticides  
• It will lower the amount of stoop labor farm workers will need to perform.  
• AI will lower the percentage of the world’s working population that must farm  
• It will provide any farmer with internet access and instantaneous answers to quests such as when, what and how to plant.  

References  

1. Understanding Artificial Intelligence: What It Is and How It Is Used in Agriculture: https://edis.ifas.ufl.edu/publication/AE589  
2. The Role Of AI In Creating A More Sustainable Food System: https://www.forbes.com/sites/forbestechcouncil/2023/07/20/the-role-of-ai-in-creating-a-more-sustainable-food-system/?sh=15a2e2a75315  
3. Machine learning, explained: https://mitsloan.mit.edu/ideas-made-to-matter/machine-learning-explained  
4. AI meets agriculture with new farm machines to kill weeds and harvest crops: https://www.youtube.com/watch?v=eDUu48YCUy4 
5. What is Generative AI?: https://www.nvidia.com/en-us/glossary/data-science/generative-ai/#:~:text=Generative%20AI%20models%20can%20take,or%20turn%20video%20into%20text  
6. Norm: https://www.fbn.com/norm  
7. Farmers Are Tentatively Embracing AI: https://ambrook.com/research/technology/ChatGPT-AI-farming-hallucinations
8. Improving the Speed and Efficiency of Agricultural Extension with AI: https://farmerchat.digitalgreen.org/  
9. American Agriculture’s Dependence on Immigrant Worker: https://ualr.edu/socialchange/2021/03/28/american-agricultures-dependence-on-immigrant-workers/  
10. How John Deere plans to build a world of fully autonomous farming by 2030: https://www.cnbc.com/2022/10/02/how-deere-plans-to-build-a-world-of-fully-autonomous-farming-by-2030.html

Glossary  

• Artificial Intelligence (AI):  A method that allows computer systems to think in a manner approximating that of the human mind  
• Generative AI allows users to make inquiries based on a wide variety of inputs and provide outputs, including animation, sound or images. Generative AI models are available targeting agricultural issues.  
• Machine Learning: The discipline that enables computers to learn without explicitly being programmed.   
• Stoop Labor: Farm work that must be done with the person stooping or squatting. Extremely bad for the laborer’s health.