Predictive Geological Sketching for Consistent Digital Interpretation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Geologists' sketches of geological data can vary significantly due to individual drawing abilities, making harmonization difficult, and existing digital tools disrupt the intuitive drawing process, slowing data recording.
Innovation Solution
A computer-implemented method using machine learning to predict geological features from user sketches, incorporating training with diverse user groups to enhance accuracy and efficiency in digital representation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional well logging methods are used, then basic formation data can be obtained, but the data lacks the predictive capability needed for accurate reservoir characterization and drilling risk assessment
Solution Approach 1:
The system performs predictive analysis before drilling operations by integrating geological data, well log data, and core data to generate predicted lithology, porosity, and permeability values. This preliminary action allows drilling risks to be assessed and mitigated before they occur, rather than reacting to problems during drilling.
Solution Approach 2:
A computer system with processor and memory acts as an intermediary between raw well log data and geological interpretation. The system uses trained neural networks and predictive algorithms to process raw data and generate meaningful predictions about formation properties, bridging the gap between measurement and understanding.
2Manufacturing precision
If extensive well logging and analysis are performed, then more accurate formation data can be obtained, but time and cost increase significantly
Solution Approach 1:
The system replaces manual geological analysis with automated computer-based analysis using neural networks and predictive algorithms. This substitution of mechanical/manual processes with computational methods maintains high accuracy in formation characterization while dramatically reducing the time required for analysis.
Solution Approach 2:
The system transforms raw well log parameters into predicted formation properties (lithology, porosity, permeability) through computational processing. By changing the parameters from raw measurements to interpreted predictions, the system provides actionable insights faster than traditional methods.
3Reliability
If traditional well logging approaches are used, then standard formation data is collected, but drilling risks cannot be effectively predicted or mitigated
Solution Approach 1:
The system performs multiple functions within a single integrated platform: it processes well log data, core data, and geological data; generates predicted lithology, porosity, and permeability; assesses drilling risks; and provides recommendations. This multi-functionality improves reliability by considering multiple factors simultaneously while managing complexity through integration.
Solution Approach 2:
The system uses trained neural networks that learn from historical data and provide feedback on formation properties. The predictive algorithms continuously refine their accuracy by comparing predictions with actual formation data, improving reliability over time while the automated feedback loop manages processing complexity.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method for representing a geology includes receiving one or more drawing strokes as part of a geological feature sketch, predicting, using a machine learning model, one or more predicted geological features based at least in part on the one or more drawing strokes before the sketch is complete, displaying the one or more predicted geological features, receiving a selection of one of the one or more predicted geological features, and generating an image representing the geology including a digital representation of the selected one of the one or more predicted geological features.