Oil Field Log Image Analysis for Automated Lithology Interpretation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Oil field log interpretation relies heavily on human expertise and is not automated, and complete data sets, such as ASCII encoded data, are often unavailable, necessitating manual inspection of image data like PDF files.
Innovation Solution
An oil field log image processor utilizing a large language model (LLM) and neural networks to analyze image data from logs, convert graphical information into text-encoded data, and generate subsurface properties and operational insights.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If automated interpretation methods are implemented, then productivity and efficiency are improved, but the system becomes more complex and requires advanced technology infrastructure
Solution Approach 1:
The patent replaces manual visual inspection (mechanical human process) with an automated computer-based system using optical character recognition (OCR) and natural language processing. The system automatically extracts text from log images, processes the data, and generates interpretations, eliminating the need for manual engineering analysis while significantly improving productivity.
Solution Approach 2:
The patent introduces an intermediary processing layer that converts image data into structured text data through OCR technology. This intermediary step bridges the gap between unstructured log images and the automated interpretation system, enabling subsequent natural language processing and analysis without requiring direct human intervention.
2Measurement precision
If complete data sets including text encoded data are available, then measurement precision and analysis accuracy are improved, but data availability and accessibility are reduced
Solution Approach 1:
The patent creates a digital copy of the text information from log images using OCR technology. Instead of requiring access to the original text-encoded data files, the system extracts and replicates the textual content from visual representations, making the information accessible and processable even when only image formats are available.
Solution Approach 2:
The patent transforms the data format parameter from image format to text format through OCR processing. This parameter change enables the system to work with standard text-based analysis methods and natural language processing techniques, improving both accessibility and precision simultaneously by converting between different data representations.
3Ease of operation
If manual visual inspection is performed, then ease of operation is maintained, but loss of time and productivity are increased
Solution Approach 1:
The patent performs preliminary automated processing of log data before human review is needed. The system pre-extracts text from images, pre-processes the data structure, and pre-generates initial interpretations, so that when engineers do review the logs, they are working with already-processed information rather than raw images, significantly reducing their time investment while maintaining ease of operation.
Data Source
AI summary
Some implementations include a computer-implemented method for determining one or more subsurface properties from one or more graphical oil field logs created during oil field of one or more oil field operation. The computer-implemented method may include graphically analyzing the graphical oil field logs; determining one or more regions of interest in the graphical oil field logs; generating, based on the graphical analysis of the graphical oil field logs, a respective textual description for each respective region of interest in the graphical oil field logs; and determining, by a large language model (LLM), one or more lithologies for each respective region of interest based on the textual description for the respective region of interest.