Field robotics
BroccoliBot
A computer-vision-guided robotic system designed to identify and mechanically harvest broccoli in the field.
Overview
BroccoliBot brought sensing, motion, and a physical cutting system together in a single autonomous harvesting concept. An RGB-depth camera and image processing located viable plants, while steering controls and adjustable blades handled the physical harvest.
- Robotics
- Computer vision
- Mechatronics
- Field testing
The challenge
Broccoli varies in height, spacing, and appearance, and the system needed to work outside a controlled lab. The useful solution had to perceive a plant, position the machine, and cut consistently in real field conditions.
The approach
The team combined field research, computer modeling, component fabrication, code, iterative prototyping, and testing. The resulting system joined visual detection with mechanical steering and a height-adjustable cutter.
The outcome
The work was awarded U.S. Patent 11,343,967 on May 31, 2022, recognizing a practical approach to robotic automation of mechanical field harvesting.
Role & attribution
Abraham was a member of the Crescent Valley student robotics team and is a named co-inventor on the patent. Public sources describe the team scope; individual responsibilities will be expanded here after confirmation.
Public record