Robotic Harvester Selective Crown Harvesting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for harvesting vegetables like cauliflower and broccoli are labor-intensive, inefficient, and prone to damage due to manual handling and lack of automation in determining readiness and separating edible from non-edible portions, leading to inefficiencies and waste.
Innovation Solution
A robotic harvester system equipped with imaging components, machine learning algorithms, and robotic arms that image broccoli plants to determine readiness and selectively harvest the edible crowns, using sensors and actuators to grip and cut the plants while minimizing damage and waste.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual harvesting is used, then workers can visually inspect each plant to determine readiness, but the process becomes labor-intensive and inefficient
Solution Approach 1:
The patent replaces the manual visual inspection system with an automated imaging and machine learning system. Imaging components capture images of broccoli plants, and machine learning algorithms automatically determine readiness for harvesting, eliminating the need for manual visual inspection while maintaining accuracy and significantly improving harvesting efficiency.
Solution Approach 2:
The system enables the harvesting process to determine plant readiness autonomously through automated imaging and machine learning analysis, without requiring human workers to perform visual inspections. The system self-evaluates plant maturity based on captured images and automatically makes harvesting decisions.
2Ease of operation
If manual handling stages are used, then workers can sort and process plants, but bruising and damage occur during handling
Solution Approach 1:
The patent replaces manual handling and sorting operations with automated robotic systems. Robotic arms with specialized end effectors harvest plants directly from the field and transfer them to collection containers, eliminating multiple manual handling stages that cause bruising and damage while maintaining operational capability.
Solution Approach 2:
The system performs preliminary sorting and identification of harvestable plants through imaging and machine learning before physical contact is made. This allows the robotic system to prepare for precise harvesting actions that minimize handling and reduce damage to the plants.
3Productivity
If conventional harvesting methods are used, then all plants are harvested, but this includes non-edible portions and leads to waste
Solution Approach 1:
The patent replaces conventional bulk harvesting methods with an automated system that uses imaging components and machine learning to identify and selectively harvest only the edible portions of plants. The system distinguishes between harvestable broccoli heads and non-edible portions, harvesting only the valuable parts and leaving the rest in the field, thereby reducing waste while maintaining high harvesting speed.
Solution Approach 2:
The system extracts and harvests only the specific edible portions (broccoli heads) from the plants, separating them from non-edible portions such as stems and leaves. This selective extraction approach ensures that only valuable harvestable material is collected, minimizing waste and maximizing resource efficiency.
4Extent of automation
If automated harvesting systems are attempted, then labor is reduced, but challenges remain in observing harvestable portions and gripping plants due to varying plant nature and field conditions
Solution Approach 1:
The patent employs imaging components capable of capturing images under varying field conditions (different lighting, weather, plant sizes, and orientations). The machine learning algorithms are trained to recognize and adapt to the varying nature of plants and field conditions, enabling the automated system to reliably detect and measure harvestable portions despite environmental variations.
Solution Approach 2:
The robotic arms and end effectors are designed with dynamic capabilities to adapt to varying plant positions, sizes, and orientations. The system can adjust its gripping force and positioning in real-time based on feedback from the imaging system, allowing it to handle the varying nature of plants and field conditions effectively while maintaining high automation levels.
Data Source
AI summary
A harvester travels along a route within a field and selectively harvests edible crowns ready for harvesting. As the harvester travels along the route, a position of the harvester and/or the edible crowns may be determined using an encoder, imaging device(s), and/or navigational system(s). For example, image(s) captured by the imaging device(s) may be used to determine a location of the edible crowns and/or global positioning satellite (GPS) coordinates may indicate a location of the harvester within the field. These locations may be used for instructing harvesting components to harvest the edible crowns. For example, the harvester may include robotic arms having end effectors that harvest the edible crowns. Knowing the location of the edible crowns and/or the harvester therefore allows for the accurate placement of the end effectors for harvesting the edible crowns.