Article Stacking Placement Learning for Mixed-Size Loads
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing article stacking systems struggle to efficiently stack articles of varying sizes and shapes without pre-defined patterns, especially when information about the articles is not available in advance, leading to inefficiencies and instability in the stacking process.
Innovation Solution
A machine learning apparatus that determines optimal placement for articles based on their properties and site situations by learning from data collected during stacking operations, using a state observation section, label data acquisition section, and learning section to calculate and update a correlation model for accurate placement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If articles are stacked according to a preset pattern, then the stacking process is simple and fast, but the stacking efficiency and stability deteriorate when articles have different sizes, shapes, or properties
Solution Approach 1:
The patent implements dynamic stacking pattern generation that adapts to different article properties in real-time. The system transitions from static preset patterns to dynamic patterns that are generated based on article characteristics such as size, shape, and weight, allowing the stacking configuration to change flexibly according to the specific articles being processed
Solution Approach 2:
The system changes stacking parameters (position, orientation, layer configuration) based on article properties. By adjusting these parameters dynamically according to article characteristics, the system optimizes stacking efficiency and stability for different types of articles without requiring manual pattern reconfiguration
2Productivity
If articles are stacked without pre-defined patterns to improve loading efficiency, then loading efficiency improves, but the difficulty of determining optimal placement increases
Solution Approach 1:
The system employs self-service through automated machine learning models that independently determine optimal placement positions. The machine learning apparatus automatically analyzes article properties and generates placement decisions without human intervention, enabling the system to improve loading efficiency while managing placement determination complexity through automation
Solution Approach 2:
The patent replaces manual or rule-based placement determination mechanisms with machine learning-based intelligent systems. This substitution enables the system to handle complex placement decisions by leveraging pattern recognition and predictive algorithms, thereby improving loading efficiency without proportionally increasing operational complexity
3Adaptability or versatility
If article information is not obtained in advance, then the system is more flexible, but the ability to predict optimal stacking positions deteriorates
Solution Approach 1:
The system performs preliminary actions by pre-training machine learning models using historical article data and stacking patterns. This preliminary learning phase enables the system to develop predictive capabilities that can accurately determine optimal placement positions even when specific article information is not available in advance, maintaining both flexibility and prediction accuracy
Solution Approach 2:
The system implements feedback mechanisms where stacking outcomes are continuously monitored and used to refine machine learning models. This feedback loop allows the system to improve its placement prediction accuracy over time while maintaining flexibility in handling diverse article types, as the models adapt based on actual stacking performance data
Data Source
AI summary
A machine learning apparatus of an article stacking apparatus observes, as a state variable representing the current environmental state, stacking status data indicating the stacking status of an article stacking area and article information data indicating information on an article to be stacked, and acquires, as label data, article placement data indicating a placement of the article in the stacking area. The machine learning apparatus learns article placement data in association with the stacking status data and the article information data using the state variable and the label data.