Robotic Fruit Picking With Vision-Guided Arms for Table Top Farms
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current robotic fruit picking systems are expensive, require significant changes in farming practices, and are not compatible with European table top growing systems, leading to high production costs and labor market fluctuations, making them uncompetitive with human labor.
Innovation Solution
A robotic fruit picking system utilizing a tracked rover with a 3D stereo camera and reinforcement learning for efficient navigation and picking, a picking head capable of cutting and gripping fruits, and a quality control subsystem for grading, designed to work autonomously and collaboratively with human pickers, using lower-cost off-the-shelf components and state-of-the-art computer vision techniques.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If sophisticated hardware and naive robot control systems are used for robotic fruit picking, then the robot can perform picking operations, but the system becomes expensive and requires carefully controlled environments, reducing commercial success
Solution Approach 1:
The patent replaces sophisticated mechanical hardware systems with software-based solutions. Specifically, it uses advanced computer vision algorithms and machine learning models to perform fruit detection, localization, and picking decisions, substituting complex mechanical sensing and control systems with computational approaches that run on standard computing hardware.
Solution Approach 2:
The patent changes the operational parameters of the picking system by using reinforcement learning to adapt picking strategies dynamically. The system learns optimal picking parameters (force, speed, grip position) through trial and error, allowing it to perform effectively without requiring precisely controlled environmental parameters or expensive specialized hardware.
2Productivity
If robotic picking systems are designed with high production capacity, then harvesting efficiency improves, but the cost per unit picking capacity increases disproportionately compared to small autonomous machines
Solution Approach 1:
The patent divides the harvesting task into smaller sub-tasks performed by multiple independent picking heads. Each picking head is a relatively simple autonomous unit that can detect, localize, and pick fruit independently. By segmenting the overall system into multiple smaller units rather than one large complex machine, the system achieves high total productivity while keeping individual unit costs low and manageable.
Solution Approach 2:
The picking heads are designed to be universal and adaptable to different fruit types and growing conditions. Rather than designing specialized large machines for specific high-volume applications, the system uses multi-functional picking heads that can be deployed in various configurations and environments, reducing the need for expensive application-specific hardware.
3Manufacturing precision
If robotic systems require carefully controlled environments to operate, then picking precision improves, but adaptability to different growing systems (such as European table top systems) decreases
Solution Approach 1:
The patent implements continuous feedback loops using computer vision systems that monitor fruit position, orientation, and ripeness in real-time. This visual feedback allows the robot to adjust its picking actions dynamically, maintaining high precision without requiring a controlled environment. The system adapts to different growing configurations (including table top systems) by continuously sensing and responding to the actual fruit and plant positions rather than relying on pre-programmed environmental assumptions.
Solution Approach 2:
The system uses dynamic control strategies where picking parameters (grip force, arm speed, approach angle) are adjusted in real-time based on visual feedback and reinforcement learning policies. This dynamic adaptation allows the robot to maintain precision across varying environmental conditions and different growing systems, including European table top configurations, without requiring static environmental controls.
Data Source
Figure 1(a)~1(b)
Figure 2
Figure 3
AI summary
A robotic fruit picking system includes an autonomous robot that includes a positioning subsystem that enables autonomous positioning of the robot using a computer vision guidance system. The robot also includes at least one picking arm and at least one picking head, or other type of end effector, mounted on each picking arm to either cut a stem or branch for a specific fruit or bunch of fruits or pluck that fruit or bunch. A computer vision subsystem analyses images of the fruit to be picked or stored and a control subsystem is programmed with or learns picking strategies using machine learning techniques. A quality control (QC) subsystem monitors the quality of fruit and grades that fruit according to size and/or quality. The robot has a storage subsystem for storing fruit in containers for storage or transportation, or in punnets for retail.