Predictive Maintenance Using Pretrained Feature Extraction in Manufacturing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Industrial machine failures in manufacturing processes lead to production disruptions and resource losses due to inadequate predictive maintenance, as existing methods rely solely on physical observations of machines without integrating measurement data from processed parts.
Innovation Solution
A computer-implemented method using machine learning models to generate predictive maintenance by encoding measurement data into latent representations, combining it with machine observation data to forecast maintenance needs for the next process period, thereby anticipating potential failures and reducing downtime.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing predictive maintenance methods rely solely on physical observations of machines, then the system is simple to implement, but the prediction accuracy is insufficient leading to machine failures
Solution Approach 1:
The patent merges two previously separate data sources - machine observation data (physical observations) and part measurement data (from sensors at each station) - into a unified predictive maintenance system. This combination allows the system to leverage both traditional machine condition monitoring and new product quality measurements, thereby improving prediction accuracy while maintaining reasonable system complexity through integrated processing.
Solution Approach 2:
The patent introduces machine learning models as intermediary components that process and integrate the combined data from machine observations and part measurements. These models act as mediators that transform raw data from multiple sources into actionable maintenance predictions, enabling the system to handle the complexity of multi-source data integration while delivering improved prediction accuracy.
2Productivity
If machine failures are not predicted accurately, then maintenance can be performed on schedule, but production disruptions and resource losses occur
Solution Approach 1:
The patent implements preliminary action by predicting machine failures before they actually occur. The system analyzes current machine observation data and part measurement data to forecast future maintenance needs, enabling maintenance teams to perform repairs proactively during planned downtime rather than reactively after failures disrupt production. This preliminary intervention maintains production continuity while ensuring machine reliability.
3Measurement precision
If only machine observation data is used for maintenance predictions, then data collection is simple, but the predictions lack accuracy for preventing failures
Solution Approach 1:
The patent merges two previously separate data sources - machine observation data (physical observations) and part measurement data (from sensors at each station) - into a unified predictive maintenance system. This combination allows the system to leverage both traditional machine condition monitoring and new product quality measurements, thereby improving prediction accuracy while maintaining reasonable system complexity through integrated processing.
Data Source
AI summary
A computer-implemented system and method includes establishing a station sequence that a given part traverses. Each station includes a machine that performs at least one operation with respect to the given part. Measurement data, which relates to attributes of a plurality of parts that traversed the plurality of machines, is received. The measurement data is obtained by sensors and corresponds to a current process period. A first machine learning model is pretrained to generate (i) latent representations based on the measurement data and (ii) machine states based on the latent representations. Machine observation data, which relates to the current process period, is received. Aggregated data is generated based on the measurement data and the machine observation data. A second machine learning model generates a maintenance prediction based on the aggregated data. The maintenance prediction corresponds to a next process period.