Robotic Broccoli Crown Harvesting With AI Ripeness Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for harvesting crops like cauliflower and broccoli are labor-intensive, inefficient, and prone to bruising or damage, with automated systems facing challenges in identifying and separating edible portions from non-edible portions due to varying plant conditions.
Innovation Solution
A harvester equipped with imaging systems, robotic arms, and machine learning algorithms to detect and selectively harvest edible crowns by analyzing plant characteristics, removing leaves, and cutting the crowns from the stem, while navigating through fields using GPS and de-leafing components.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual harvesting is used, then workers can visually inspect and selectively harvest ripe plants, but the process is labor-intensive and inefficient
Solution Approach 1:
The patent replaces manual visual inspection and harvesting operations with an automated system using imaging devices (optical systems) to detect ripe plants and robotic mechanisms to harvest them. The imaging devices capture images of plants, and a processing system automatically identifies ripe plants based on image analysis, eliminating the need for manual visual inspection by workers.
Solution Approach 2:
The system enables the harvesting machine to autonomously identify and harvest ripe plants without continuous human intervention. The processing system automatically analyzes images, determines which plants are ripe, controls the harvesting mechanisms to harvest identified plants, and can adjust operating parameters based on real-time detection, making the system self-sufficient in the harvesting decision-making process.
2Reliability
If conventional harvesting techniques are used, then multiple handling stages occur, but this gives rise to bruising or damage
Solution Approach 1:
The patent segments the harvesting function into distinct operational phases: detection phase (imaging devices capture plant images), identification phase (processing system analyzes images to identify ripe plants), and execution phase (harvesting mechanisms selectively harvest identified plants). This segmentation allows each phase to be optimized independently, reducing unnecessary handling and minimizing damage to crops.
Solution Approach 2:
The imaging devices and processing system serve as intermediaries between the harvester and the plants. Instead of direct manual contact with all plants, the system uses optical detection as an intermediary to identify target plants, and controlled robotic mechanisms as intermediaries to perform the actual harvesting, reducing unnecessary contact and handling of non-target plants.
3Productivity
If the harvester moves at high speed, then productivity increases, but positioning accuracy for selective harvesting deteriorates
Solution Approach 1:
The system performs preliminary imaging and plant identification before the harvesting execution phase. The imaging devices capture images of upcoming plants in advance, the processing system analyzes these images to identify ripe plants, and pre-calculates harvesting positions. This preliminary action allows the system to maintain high forward speed while ensuring accurate positioning for each harvesting operation.
Solution Approach 2:
The system uses real-time feedback from imaging devices to continuously monitor plant positions and ripeness status. The processing system compares detected plant positions with planned harvesting positions, and dynamically adjusts the harvesting mechanisms' positioning and timing. This closed-loop feedback control enables accurate selective harvesting even while the machine moves at high speed through the field.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A harvester that determines whether edible crowns are ready to be harvested and selectively harvests the edible crowns that are ready for harvesting. The harvester may include sensors, such as an imaging system, for detecting the edible crowns of individual broccoli plants. Image data from the imaging system may be provided as an input to a machine-learning model to determine a maturity (or immaturity) of the edible crowns. If the edible crowns are ready for harvesting, mechanical pickers harvest the edible crowns. For example, the harvester may include robotic arms having end effectors that cut the edible crowns from a remainder of the broccoli plant. The harvester may be configured to continuously harvest the edible crowns as the harvester moves about a field. In some instances, the harvester may include any number of robotic arms for harvesting the edible crowns across multiple rows of broccoli plants.