PLC Fault Detection Using Buffered Process Images and ML
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional manufacturing and automation systems lack effective fault detection and monitoring capabilities, making it difficult for operators to diagnose the root cause of faults in automation devices, leading to increased downtime and costs due to inadequate data collection and analysis.
Innovation Solution
An automation device method and system that includes storing input and output process images in a buffer, using a ring buffer to collect data, and employing a machine learning model to predict faults, allowing for improved data collection and analysis for fault diagnosis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional error code collection methods are used, then fault detection is simple, but diagnostic capability is insufficient for operators to identify root causes
Solution Approach 1:
The system performs preliminary data collection and processing by capturing process images, sensor data, and operational parameters before faults occur. This preparatory action ensures that comprehensive diagnostic information is already available when a fault happens, eliminating the need for post-fault data gathering and enabling immediate root cause analysis.
Solution Approach 2:
The patent introduces an intermediary data processing layer that collects and correlates information from multiple sources including process images, sensor readings, and operational parameters. This intermediary system bridges the gap between raw data and diagnostic insights, providing operators with synthesized fault analysis results rather than just raw error codes.
2Loss of information
If detailed data collection is implemented for fault analysis, then diagnostic capability improves, but system complexity and data processing requirements increase
Solution Approach 1:
The data collection system is segmented into modular components: process image capture modules, sensor data acquisition modules, operational parameter collection modules, and analysis modules. Each segment handles specific types of data independently, which can be configured and maintained separately, reducing overall system complexity while maintaining comprehensive data collection capabilities.
Solution Approach 2:
The system incorporates automated analysis capabilities that process collected data and generate diagnostic insights without requiring complex external processing. The automated fault detection and analysis features enable the system to self-diagnose issues, reducing the burden on operators and simplifying the overall data processing architecture.
3Loss of information
If comprehensive monitoring data is collected, then root cause analysis capability improves, but downtime for data processing and maintenance increases
Solution Approach 1:
The system continuously collects and pre-processes monitoring data during normal operation, preparing diagnostic information in advance before faults occur. This preliminary data preparation ensures that when a fault happens, the analysis can be performed immediately using pre-processed information, minimizing downtime while maintaining comprehensive diagnostic capabilities.
Solution Approach 2:
The patent replaces manual data analysis and fault diagnosis procedures with automated electronic data processing and analysis systems. This substitution enables rapid processing of comprehensive monitoring data without requiring extended manual intervention, significantly reducing maintenance downtime while maintaining thorough diagnostic analysis.
Data Source
Figure 1~2
Figure 3~4
Figure 5
AI summary
Provided are systems, methods and devices for automation devices. The method includes providing an automation device for controlling at least one piece of automation equipment; receiving, at an input of the automation device, input data; storing, at an input process image, the input data; executing, at a processor of the automation device, an automation program to generate output data based on the input data; storing, at an output process image of the automation device, the output data; and storing, in a buffer, the input process image and the output process image.