Key Takeaways
- NC State University researchers are developing robotic systems to assist with labor-intensive vegetable production tasks.
- The project focuses on automating staking, monitoring, and harvesting in crops like tomatoes and peppers.
- AI, computer vision, LiDAR, and robotics are central to the system design.
- The initiative aims to address labor shortages and reduce physical strain on farm workers.
- Research is supported by state and federal funding, including the USDA.
NC State Explores Robotics for Vegetable Crop Management
Researchers at North Carolina State University are developing a suite of AI-enabled robotic tools designed to assist with labor-intensive tasks in vegetable production. The work is led by faculty affiliated with the N.C. Plant Sciences Initiative (N.C. PSI) and focuses on crops such as tomatoes and peppers, which still rely heavily on manual labor for staking, monitoring, and harvesting.
While mechanization is well established in row crops like corn and soybeans, many specialty crops require hands-on management. According to the research team, this dependence on manual labor is becoming increasingly challenging due to workforce availability and rising costs.
AI and Sensors Enable Field Navigation and Precision Tasks
Robots Designed for Real-World Farm Conditions at NC State
One of the primary robotic platforms under development is a self-driving system equipped with stereo cameras and LiDAR sensors, allowing it to map fields and navigate uneven terrain. The platform can carry multiple tools, including a robotic hammer designed to drive stakes into the ground at precise intervals without damaging plants.
“It’s similar to how autonomous household robots map indoor spaces,” said Emmanuel Torres, professor of horticultural sciences, who is collaborating on the project.
Staking is a physically demanding and repetitive task that requires thousands of stakes and multiple passes through a field as plants grow. Automating this process could reduce labor demands while maintaining consistency in plant support systems.
Building the Data Foundation for Agricultural AI
To train AI systems to distinguish crops from weeds and other field objects, the researchers developed a separate imaging tool equipped with high-resolution cameras. Mounted on a tractor, the system captures large volumes of top-down and side-view images of plants.
Supported by funding from the N.C. General Assembly and the U.S. Department of Agriculture, the tool collected approximately 50,000 images during initial field trials. These datasets are being used to improve machine vision models that could eventually support automated crop monitoring for pests, diseases, and nutrient deficiencies.
Research Into Robotic Tomato Harvesting
Students Test AI-Guided Harvesting Concepts
In parallel, engineering students are working on prototypes for robotic tomato harvesting. These systems combine depth-sensing cameras, AI-based fruit detection, and soft silicone grippers designed to handle tomatoes without causing damage.
The researchers acknowledge that human workers are currently faster and more efficient, but note that machines could operate continuously and complement existing labor rather than fully replace it.
Additional Tools Under Development at NC State
Other concepts include a device designed to collect insects from crop canopies to support pest identification and integrated pest management research. The team expects to complete an initial prototype in early 2026.
