AI Geological Data Processing Tool for Secure Subterranean Feature Identification

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Simulating complex reservoir geometries and behaviors using computer models is challenging due to high-contrast features like fractures and faults, and training AI for geological data processing is time-consuming and proprietary data-dependent, limiting its effectiveness in petroleum extraction.

Innovation Solution

A method involving a client system with AI tools, such as convolutional neural networks, that allows for training on proprietary data without exposing it to the provider system, enabling identification of geological features and modification of parameters for improved subterranean volume modeling.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If AI is trained using proprietary training data, then the accuracy of geological feature identification is improved, but the training data security is compromised due to exposure to provider system

Engineering Contradiction:
Improveaccuracy of geological feature identificationVSAvoidtraining data security
Core Design Contradiction:
Measurement precisionVSObject-affected harmful factors

Solution Approach 1:

The patent introduces an intermediary mechanism where the client system trains the AI model locally using their proprietary training data without exposing it to the provider system. The trained AI model is then transferred to the provider system for deployment, thus mediating between the need for accurate training and data security requirements

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent segments the AI development process into two distinct phases: training phase performed by the client system using proprietary data, and deployment phase performed by the provider system. This segmentation allows each party to operate within their security boundaries while achieving the overall goal

Inventive Principle:
Principle #1Segmentation

2Ease of operation

If AI training is performed externally by provider system, then the process is simplified for clients, but the training data must be exposed to provider system creating security risks

Engineering Contradiction:
Improvesimplicity of AI training process for clientsVSAvoidtraining data exposure
Core Design Contradiction:
Ease of operationVSObject-affected harmful factors

Solution Approach 1:

The client system acts as an intermediary that performs the training operation locally, eliminating the need for clients to directly expose their data to the provider system while still achieving simplified operation through automated local training processes

Inventive Principle:
Principle #24Intermediary (Mediator)

3Object-affected harmful factors

If training data is protected from provider system, then data security is improved, but the provider system cannot access training data for model improvement

Engineering Contradiction:
Improvetraining data protectionVSAvoidmodel improvement capability
Core Design Contradiction:
Object-affected harmful factorsVSAdaptability or versatility

Solution Approach 1:

The patent uses copying by transferring the trained AI model from the client system to the provider system. The provider system receives a copy of the trained model without accessing the proprietary training data, thus protecting data security while enabling model deployment and potential future improvements through other means

Inventive Principle:
Principle #26Copying

Data Source

PatentEP3535607B1Seismic data processing artificial intelligence
Publication Date: 2022.04.20 SERVICES PETROLIERS SCHLUMBERGER SA
  • EP3535607B1 patent drawingFigure 1
  • EP3535607B1 patent drawingFigure 2
  • EP3535607B1 patent drawingFigure 3

AI summary

Systems, methods, and non-transitory computer-readable media for processing geological data, of which the method includes receiving a geological data processing tool at a client system. The geological data processing tool includes artificial intelligence, and the geological data processing tool is generated by a geological processing tool provider. The method also includes obtaining training data for the geological data processing tool. The training data includes a plurality of labels. The method also includes training the geological data processing tool based on the training data, receiving data representing a physical, subterranean volume, identifying one or more geological features in the subterranean volume by using the geological data processing tool after training the geological data processing tool, and modifying, using the client system, one or more parameters of the geological data processing tool, or one or more labels of the plurality of labels.