OCR Screenshot Capture for Post-Mortem Analysis of Tripped Field Devices
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Operators in process industries face challenges in performing post-mortem analysis of tripped field devices due to the lack of comprehensive tools for reviewing past information and configuring critical devices, especially when process units halt, as existing systems do not allow for easy retrieval or analysis of previous critical process parameter values.
Innovation Solution
The implementation of Optical Character Recognition (OCR) and Intelligent Character Recognition (ICR) techniques within the Field Asset Maintenance System (FAMS) application to capture and analyze screenshots of operator console displays, allowing for the storage and reconstruction of past process unit states, enabling operators to perform root cause analysis of tripped devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If operators manually review console station displays to diagnose tripping issues, then they can identify currently tripped devices, but they cannot access past information for post-mortem analysis
Solution Approach 1:
The system performs preliminary actions by capturing screenshots and extracting process parameter values before tripping events occur. The continuous monitoring and data extraction mechanism ensures that historical information is preserved and made available for later post-mortem analysis, eliminating the loss of critical diagnostic information.
Solution Approach 2:
The system creates visual copies of the console station display through screenshot capture. These screenshots serve as reproducible records that can be analyzed later without requiring access to the original live display system, enabling operators to review past states and perform comprehensive post-mortem analysis.
2Reliability
If the system captures and stores detailed screenshot information for post-mortem analysis, then operators can perform comprehensive root cause analysis, but the system complexity and data processing requirements increase
Solution Approach 1:
The system extracts only the essential diagnostic information from screenshots, such as process parameter values, device states, and temporal information. By selectively extracting relevant data rather than processing entire image files, the system maintains high diagnostic reliability while reducing the complexity of data storage and processing requirements.
Solution Approach 2:
The system introduces an intermediary processing layer that bridges the visual display information and the diagnostic analysis requirements. This intermediary extracts structured data from unstructured screenshot images, transforming them into analyzable formats without requiring complex direct processing of raw image data, thus balancing reliability with system simplicity.
3Loss of time
If the system continuously monitors and captures screenshots at frequent intervals, then operators can access detailed temporal information about parameter transitions, but the data storage and processing load increases
Solution Approach 1:
The system applies local quality by capturing screenshots at varying intervals based on the specific monitoring needs of different process parameters and operational conditions. Critical parameters experiencing rapid changes are monitored more frequently, while stable parameters are checked less often, optimizing temporal resolution while minimizing overall data storage requirements.
Data Source
Figure 1
Figure 2A
Figure 2B
AI summary
A method includes capturing (208) at least one screenshot of a display screen including an initial screenshot (300). The method includes removing (212) text from the initial screenshot to generate a base image (500, 700). The method includes identifying (216) a background (Region 0) of the initial screenshot as a closed region. The method includes, for each of the at least one screenshots: storing (222) a time (970) of capturing the screenshot; identifying (212, 226) text, text color, and text location in the screenshot; identifying (216, 230) each closed region (Regions 1-12) in the screenshot that is different from the background of the initial screenshot, and a region color and region location for each identified closed region in the screenshot; storing (222) the region color and the region location for each identified closed region; and storing (214, 228) the text color and the text location of the identified text.