Distributed Equipment Failure Risk Detection in Industrial Processes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In industrial processes like semiconductor manufacturing and chemical processes, equipment failure risk analysis is hindered by the need to process large volumes of data, which takes a long time due to extensive operational conditions and features recorded over long durations, necessitating a more efficient method for predictive maintenance.
Innovation Solution
A computer-implemented method and system that distributes equipment operations data across a cluster of servers using a distributed file system, performing parallel distributed processing with MapReduce operations to compute and aggregate operation features, constructing a target table for risk failure analysis, and predicting equipment lifespan and failure probabilities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional sequential processing is used to analyze equipment operations data, then comprehensive failure risk analysis can be performed, but the processing time becomes excessively long
Solution Approach 1:
The patent segments the large equipment operations data set into smaller partitions that can be distributed across multiple servers in a cluster. Each server processes a specific partition independently, allowing parallel computation of operation features. This segmentation enables the system to maintain comprehensive analysis coverage while significantly reducing overall processing time through distributed parallel processing.
Solution Approach 2:
The patent merges the processing capabilities of multiple servers into a unified distributed computing system. By combining the computational resources of multiple servers working in parallel on different data partitions, the system achieves faster overall processing while maintaining the comprehensiveness of failure risk analysis. The aggregation of operation features from multiple servers produces the complete analysis result.
2Measurement precision
If all operational conditions and features are recorded throughout equipment operation, then accurate failure risk prediction is achieved, but data volume becomes unmanageably large
Solution Approach 1:
The patent extracts only the essential operation features from the massive raw operational data. Instead of processing all recorded operational conditions, the system identifies and extracts key features that are most relevant to equipment failure prediction. This extraction process maintains prediction accuracy by focusing on critical indicators while dramatically reducing the volume of data that needs to be stored and processed.
Solution Approach 2:
The patent inverts the traditional approach by first defining the essential features needed for failure prediction, then selectively collecting and processing only those features rather than attempting to process all available operational data. This inversion allows the system to achieve accurate failure risk prediction with a manageable data volume by focusing computational resources on the most predictive features.
Data Source
AI summary
Detecting equipment failure risk in industrial process may include distributing equipment operations data to a cluster of servers based on a range of time and operation specified in maintenance data associated with the equipment. From a record entry in the maintenance data, an operation and installation and maintenance time may be determined. A plurality of servers storing equipment operations data associated with the operation during a time range between the installation and the maintenance time are selected. A distributed processing operation in each of the plurality of servers is executed to run in parallel and computes operation features. The operation features are aggregated and added as an entry in a target table. Equipment failure risk is detected by risk failure analysis performed based on the target table. A signal may be sent to automatically adjust or correct one or more operation features.