Plowing Into the Future

The precision agriculture systems and services industry surpassed $6.6 billion in value last year, and the market is expected to grow at a compound annual growth rate of nearly 15% for the remainder of the decade. This impressive growth underscores the increasing importance of precision agriculture in modern farming practices. Precision agriculture represents a significant shift from traditional farming methods. It leverages cutting-edge technologies, such as GPS, drones, sensors, and data analytics, to optimize various aspects of farming operations, from planting and irrigation to pest control and harvest management.

While precision agriculture holds immense promise for improving crop yields, minimizing resource wastage, and reducing environmental impact, there is a significant challenge that must be addressed. Most commercial solutions available today are prohibitively expensive, placing them far out of reach for nearly all small-scale farmers. This creates a concerning disparity in access to the benefits of precision agriculture technology, with larger, wealthier farms reaping the rewards while smaller operations struggle to compete.

As of 2021, 89 percent of farms in the United States were small family farms according to the USDA. These farms account for nearly 50 percent of all farmland in the country, so excluding them from participating in cutting-edge precision agriculture techniques is a detriment not only to these small farmers, but also to food supplies and the environment.

A software engineer and hacker that goes by the handle kp finds this situation troubling, and so decided to try and do something about it. That resulted in the development of an open robotic platform called the Agrofelis Robot. Designed to be low-cost, customizable, and capable of running modern AI algorithms, this wheeled robot could help small farmers keep up with the rapid changes in their field (pun intended).

The chassis of the Agrofelis Robot was designed such that it could reasonably be manufactured in a local machine shop. The four wheels are driven by energy-efficient in-hub motors that are regulated via real-time feedback sensors. A pair of independent linear actuators support the robot’s steering capabilities. The sturdy design allows the Agrofelis Robot to carry a load of up to 550 pounds.

Power is delivered by eight-element LiFePO4 batteries rated at 302 amps, providing eight kWh of energy in a one square meter space. An intelligent battery management system was included to make the most of the available energy resources, and also to power the onboard sensors and compute hardware. An NVIDIA Jetson Nano single-board computer, a Google Coral AI accelerator, and Arduino microcontrollers are included to control the robot, process sensor data, run AI algorithms, handle communications, and more. A high-speed camera is also onboard for computer vision applications, like crop inspections.

On top of the robot, there is a rotating base where various implements can be attached to suit a variety of applications. These implements could include sprayers, fertilizer spreaders, robotic picking arms, or additional sensors.

Given the open source, programmable design of the Agrofelis Robot, the use cases for the robot can be heavily tailored to an individual farm’s needs, and can even change throughout the course of a season, as needs shift. It was also noted by kp that this flexibility might allow the robot to take on completely different tasks, like fighting wildfires.

As built in this project, the parts list came in at about $4,000, excluding shipping and assembly costs. Undoubtedly, building a custom Agrofelis Robot would take a lot of time and effort, but it may well be worth it for many small farmers considering the low cost and cutting-edge capabilities of a finished build. Be sure to check out the project write-up for the details you will need to build your own robot.