Operation Prediction Models for Skilled Task Automation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for controlling robots and plants require extensive time and effort to create programs that replicate human operations, often relying on trial and error, and struggle to express nuanced human experiences and intuitions effectively.
Innovation Solution
An operation prediction system utilizing machine learning to classify and learn from operation case data, including environmental and human operation data, allowing for the selection of appropriate models to predict human operations in various environments, thereby automating tasks efficiently and incorporating skilled techniques.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional programming methods are used to control robots, then the robot can perform desired operations, but it requires extensive time to create and adjust programs
Solution Approach 1:
The system copies human operational patterns by capturing actual human operations and using them as training data for machine learning models. Instead of programming robots with explicit instructions, the system learns by copying human behavior patterns from observed operation data, significantly reducing program creation time while maintaining operational accuracy
Solution Approach 2:
The patent replaces traditional mechanical programming systems with machine learning-based predictive systems. Instead of manually creating and adjusting control programs, the system uses learned models to predict and generate control instructions automatically, eliminating the time-consuming iterative programming process
2Adaptability or versatility
If trial and error methods are used to automate operations, then skilled human techniques can be replicated, but extensive trial and error is required
Solution Approach 1:
The system performs preliminary learning by training machine learning models on comprehensive human operation data before actual automation deployment. This preliminary training phase captures skilled techniques and intuitions in advance, allowing the automated system to perform complex tasks without requiring extensive trial and error during actual operation
Solution Approach 2:
The system copies skilled human techniques by using actual human operation data as training samples. The machine learning models learn to replicate nuanced human skills and intuitions directly from observed behavior patterns, eliminating the need for repeated trial and error to discover effective operation methods
3Ease of manufacture
If language is used to express human experience, then programming is facilitated, but nuances are not correctly expressed
Solution Approach 1:
Instead of using language to describe human experience, the system directly copies actual operation data and uses it as training input for machine learning models. This approach preserves all nuances and subtle variations in human performance without the information loss that occurs when translating experience into language-based programming instructions
Solution Approach 2:
The patent replaces language-based programming with direct data-driven machine learning. Instead of translating human experience into textual descriptions and then into code, the system uses raw operation data to train models that directly learn operational patterns, preserving all nuanced information throughout the process
Data Source
AI summary
The automatic operation system includes a plurality of learned imitation models and a model selecting unit. The learned imitation models are constructed by machine learning of operation history data, the operation history data being classified into several groups by an automatic classification system algorithm, the operation history data of each group being learned by the imitation model corresponding to the group. The operation history data include data indicating a surrounding environment and data indicating an operation of an operator in the surrounding environment. The model selecting unit selects one imitation model from several imitation models based on a result of classifying data indicating a given surrounding environment by the automatic classification algorithm of the classification system. The automatic operation system inputs data indicating the surrounding environment to the imitation model selected by the model selecting unit to predict an operation of the operator with respect to the surrounding environment.