AI-guided produce processing platform for precise cutting and efficient low-waste handling of fruits and vegetables

A smart processing system uses conveyors, imaging, AI neural networks and robotic water/air jet cutters to orient, slice or reject fruits and vegetables like strawberries, improving speed, accuracy, yield and minimizing waste.

Background

Fresh fruits and vegetables require precise cutting and sorting to meet quality and safety standards before reaching consumers. Historically, these tasks have relied on manual labor, which can be slow, inconsistent, and costly. Variations in produce size, shape, and orientation introduce challenges for maintaining high yield and low waste, especially when removing unwanted parts such as strawberry caps. There is a growing demand for automated systems that can increase throughput while ensuring uniform cuts, reducing spoilage, and improving overall processing efficiency to keep pace with large-scale production and labor shortages.

Despite advances in mechanical sorting and basic vision systems, current approaches often struggle to adapt to natural variations in produce. Many machines rely on fixed cutting paths or simple shape recognition, leading to misaligned cuts, damaged fruit, and higher rejection rates. Inconsistent lighting and surface reflections can confuse basic cameras, while rigid mechanisms may fail to gently handle delicate items. As a result, processors face ongoing issues with fruit loss, increased maintenance, and limited ability to scale operations without significant human oversight.

Technology

This system processes fruits and vegetables on a conveyor using cameras, image analysis software, and jets of water and air. As each piece moves along, its shape and position are captured in real time, and a control system adjusts its orientation before trimming. In the strawberry-focused version, a machine learning model decides whether to keep or reject each berry, computes a smooth cutting path following key points, and directs a rapid robotic cutter along that curve.

Its main advantage is the combined sensing, decision-making, and cutting in a single, synchronized system. Real-time image analysis with machine learning assesses each item’s ripeness and shape, rejecting those that do not meet quality standards. Jets of water and air position the produce precisely for uniform cuts, and the computed curved path matches each item’s natural contour, reducing waste. Together, these features enable faster processing and more consistent yields than conventional sorting and cutting lines.

Benefits

  • High-speed robotic cutting for increased throughput
  • AI-powered accuracy yielding precise cuts and consistent quality
  • Multi-point curved cutlines maximizing fruit yield and minimizing waste
  • Automated produce discrimination reducing manual sorting and labor costs
  • Integrated networking for real-time monitoring and synchronization

Applications

  • Automated strawberry decalyxing
  • Precision fruit slicing
  • Vegetable trimming automation
  • Low-waste produce processing
  • Produce quality sorting

Additional info

A synchronized processing line integrates a conveyor belt, multi-angle imaging modules with real-time image-processing software, pneumatic water/air jets for produce positioning and sectioning, and networked control units. A convolutional neural network classifies strawberries for decalyxing, rejecting unqualified samples and computing curved multi-point cutline trajectories. Outputs guide a high-speed robotic cutter, delivering fast, accurate trimming with minimal waste.

Publication

“Design and testing of an automated high-throughput computer vision guided waterjet knife strawberry calyx removal machine”

Journal of Food Engineering Volume 211, October 2017, Pages 30-38

John Lin, Maxwell Holmes, Robert Vinson, Canran Ge, Frank C. Pogoda, Luke Mahon, Ryan Gentry, Gary E. Seibel, Xin Chen, Yang Tao