Engineering the Future of Agriculture: Precision, Sustainability, and Innovation
On June 26, 2024 in All, General, IoT by Traci Browne
Picture a farm: a family operation that stretches for a few thousand acres. They have some livestock—mostly dairy cows and hens—but their focus is a mix of crops, including wheat, beans, and lentils. What do you see in your mind’s eye?
As an engineer, you may imagine drones that use artificial intelligence (AI) and machine learning (ML) to detect the optimal yield. Or, perhaps, irrigation systems controlled by sensors that measure soil moisture.
Today’s farmers seek new precision technologies to help with day-to-day operations and increase crop yield, but viable solutions can be hard to come by. That’s where engineering comes in.
The Advent of the Tech-Savvy Farmer
Apple growers in the Lake Ontario region of western New York are getting a lot of assistance from the Cornell Cooperative Extension’s Lake Ontario Fruit Team (CCE-LOF).
“We have quite a few early tech adopters,” says Craig Kahlke, team leader for CCE-LOF and specialist in fruit quality management. “Probably fifteen to twenty-five farms that we regularly work with on any number of projects; they’re always on the cutting edge and always looking at new technology.”
Kahlke shares that these growers want to stay ahead and foster the spread of technology once it is working for them. However, the ability to justify the cost with the return on investment is holding many farms back from adopting smart agriculture technology.
Andrew Nelson, a fifth-generation farmer and software engineer, runs a 7500-acre farm in eastern Washington state. He advises that when developing devices, engineers need to be able to keep costs low because farming is a very low-margin business.
“Any product made for farming needs to either increase revenue or decrease costs overall,” says Nelson. “Whenever I work with a startup, I always tell them you must make sure that you are either much cheaper than what you’re replacing, or you’re going to increase the yield.”
Nonetheless, the yield aspect is tricky. Not only do you have to prove your products can improve on what farmers are already doing, but further complicating matters are the numerous variables that influence yield—with weather conditions emerging as a particularly prominent factor. This inherent variability in agriculture can significantly compound the challenges that engineers face in this field. Nelson advises starting with cost reduction because the cost savings help fuel growth for the farmer, allowing them to invest in more technology.
Precision Innovations on the Horizon
While there is no shortage of ideas on how to augment precision agriculture through technology, and much has been written for decades on how it will bring about farming 4.0, we’re not quite there yet. But Nelson thinks we’re on the cusp, as components like sensors and imaging tools have improved and lowered prices. For example:
• Through a grant from the US Department of Agriculture (USDA), scientists and engineers at Carnegie Mellon University are assessing the viability of using an array of multiple inexpensive linear arms to reach and pick fruit.[1] The trick is to get the robot to see the fruit amongst the foliage. To that end, they are applying cameras with multiple viewpoints, deep learning, and blowers to agitate the foliage.
• In the 2022–2026 Information Technology Strategic Plan, the USDA highlights the need for mobile augmented reality to detect and monitor an area’s physical characteristics so farmers can plan optimal cropping patterns and perform tasks such as monitoring for pest infestation, thereby streamlining the crop management processes.[2]
• Researchers at Cornell University, aided by a USDA grant, are focused on smart automation and data-driven precision animal management related to feeding, health, and reproductive monitoring and management of cows.[3] Wearable sensors on ears and legs monitor the cow’s biometrics and integrate with environmental sensors to assess performance, body weight, and health status.
Overcoming Challenges
Kahlke’s fruit growers in New York are having trouble with the accuracy and scale of the solutions they are trying.
Recently, he has been working with apple growers to help improve yield through precision thinning. When apple trees are blooming, each flower can potentially produce fruit, so these many clusters may produce hundreds or even thousands of apples. However, the optimal amount farmers aim for is about 150 apples per tree. Less fruit means bigger, tastier apples; or, in other words, marketability.
To achieve the optimal yield, the farmers must chemically thin the apples on each tree. Modeling software exists to help with that, but they are working manually with averages, not precision.
Kahlke explains, “Any proposed technology must be able to measure the fruit diameter within a half a millimeter accuracy to determine which will stay, which will have to go, and if and where the farmer needs to apply another chemical.”
However, the technology needs to not only measure the fruit, but also count the fruit accurately. Kahlke says that means not counting fruit twice from both sides of the tree and overestimating the amount of fruit shielded by limbs and shading. That problem has not yet been solved.
“We’re trying to figure out what percentage of error growers will accept. Maybe now we can live with a 20 percent error rate. But three years from now, can we get down to a 10 or 5 percent margin of error?” Kahlke wonders.
Also, the data gathered must be analyzed quickly. Kahlke says that most farmers have a good Wi-Fi® connection in their homes, but if a solution means uploading gigabytes of data to be analyzed, that may take all night or longer. Farmers don’t have time for that.
“Sometimes farmers want to be able to decide within 24 hours of the measurements because they need to get their next spray on to beat the weather. If you can’t give them answers within 24 hours, it’s completely useless.”
Yet, Nelson admits that accuracy is undoubtedly a benefit, but if gaining one percent accuracy doubles the cost, it’s not worth it. “You must remember that most of us are working now without data. So, I believe having decently reliable data is better than nothing.”
Engineering a Sustainable Farm
As innovators dream up futuristic agricultural technologies in the lab, electrical and design engineers will be responsible for transforming these ideas into optimized and sustainable solutions for the field. Precision agriculture relies on an orchestra of sensors, robotics, connectivity modules, and AI modeling coming together seamlessly—and engineers are the perfect conductors. By leveraging efficiency, automation, and analytical insight principles, they weave together a high-tech web that alleviates the labor shortage while elevating yield.
The following are just some of the ways engineers can help farmers achieve these sustainable precision practices.
Energy-Efficient Farm Tech
Innovative energy-efficient technologies can significantly enhance the sustainability and productivity of modern farms.
• Design more energy-efficient sensors, robotics, communication networks, and other farm tech using lower-power components, optimized PCB layouts, voltage scaling, smart power management techniques with microcontrollers, and energy-harvesting circuits.
• Employ precision agriculture principles to target the application of energy, water, and farm inputs like fertilizers and pesticides, thereby reducing waste and overuse.
• Design low-cost, durable solar technology, batteries, and energy storage solutions.
Optimizing Robotics and AI
Robotics and AI are critical in transforming traditional farming practices into high-efficiency, precision-oriented operations.
• Create customized robotic systems for niche, precision-oriented tasks, from seeding to weeding and pest control through post-harvest handling.
• Focus AI and ML on critical impact areas like continuous assessment of crop health, predictive modeling of crop disease and yield, and precision-guided drones/robotics.
• Build centrally managed, cloud-based AI systems that allow small farms to tap into advanced models and insights by joining an agricultural data cooperative platform.
Integrating Sensors and Connectivity
Advanced sensors and robust connectivity solutions are essential for real-time data collection and decision-making on the farm.
• Engineer low-maintenance, low-cost, low-power, and durable on-plant and in-plant sensors that provide a wealth of multispectral crop data like temperature, moisture, and nutrient levels.
• Create scalable mesh networks, gateways, and protocols that enable communication between sensors, robots, and edge devices across acres of cropland.
• Allow integration into emerging 5G and LEO-satellite rural broadband connectivity to bring broadband speeds that enable high-resolution sensing and real-time analytics.
The Future of Farming
Farming is full of challenges, and Nelson, like many engineers, loves a good challenge. Someone once asked him if his operation was a hobby farm. His wry reply: “This would be a terrible hobby.”
It’s the perfect response, emphasizing the seriousness and dedication required to run a farm while highlighting the rewards of innovative problem-solving and engineering solutions.
Sources
[1] https://portal.nifa.usda.gov/web/crisprojectpages/1021747-nri-int-collab-tree-fruit-harvesting-with-arrays-of-vision-guided-linear-robot-arms.html
[2]https://www.usda.gov/sites/default/files/documents/usda_it_strategic_plan_final.pdf
[3] https://news.cornell.edu/stories/2023/05/43m-grant-develop-farm-future-tech
