Pallet Sorting with Sensor Fusion and ML-Based Grading
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing pallet sorting systems lack efficient methods for grading and sorting pallets based on physical properties and conditions, leading to inefficiencies in pallet reuse and maintenance.
Innovation Solution
A pallet sorting system utilizing a conveyor system, loading assembly, and machine learning algorithms to analyze sensor data from cameras and non-image sensors to determine optimal conveyance paths for pallets, including a lifting mechanism, overhead conveyor, and stacking assemblies to sort and stack pallets based on condition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual pallet sorting is used, then operational simplicity is maintained, but productivity and sorting precision are insufficient
Solution Approach 1:
The patent replaces manual mechanical sorting with an automated system that uses sensors to detect pallet properties, a controller to process data and determine conveyance paths, and a robotic arm to execute sorting actions. This substitution of mechanical/manual operations with automated sensing and control systems directly resolves the contradiction by dramatically increasing productivity while managing complexity through integration.
Solution Approach 2:
The sorting system performs self-assessment of pallet conditions through integrated sensors that automatically detect properties such as wood moisture content, metal detectable matter, and physical dimensions. The system serves itself by making autonomous sorting decisions without human intervention, thereby increasing productivity while the modular design keeps complexity manageable.
2Measurement precision
If comprehensive sensor analysis is implemented, then sorting precision is improved, but device complexity increases
Solution Approach 1:
The patent combines multiple different sensor types (wood moisture sensors, metal detectors, dimensional sensors) into an integrated sensing system that operates under unified control. By merging these detection functions into a coordinated system rather than separate standalone devices, the patent achieves comprehensive measurement precision while managing overall system complexity through integration and centralized control.
Solution Approach 2:
The controller serves multiple functions: it processes data from various sensor types, determines conveyance paths, controls the robotic arm, and manages the overall sorting operation. This multi-functionality allows the system to achieve high measurement precision across different pallet properties while avoiding the complexity of having separate dedicated systems for each function.
3Productivity
If automated conveyance path determination is used, then sorting efficiency is improved, but control system complexity increases
Solution Approach 1:
The controller pre-determines the conveyance path for each pallet based on sensor data analysis before the pallet reaches the sorting point. This preliminary action allows the system to prepare sorting instructions in advance, enabling high-speed automated processing without requiring complex real-time decision-making mechanisms during the actual sorting operation.
Solution Approach 2:
The controller acts as an intermediary between the sensor system and the robotic arm, translating sensor measurements into conveyance path decisions and arm movement commands. This intermediary role simplifies the overall control architecture by creating a clear information flow: sensors provide data to the controller, which processes it and generates control signals for the robotic arm, thereby managing complexity through structured mediation.
Data Source
AI summary
A method for conveying a pallet through a pallet sorting system that includes receiving a pallet at a first position. A first set of pallet data is sensed from first sensors, where the first set of pallet data is non-image data. A second set of pallet data is sensed by image sensors. A logic module receives the first set of pallet data and the second set of pallet data. A subset of pallet data is selectively communicated to a machine learning module, which determines a third set of pallet data. A conveyance path is determined based at least on the third set of pallet data.