Automated Drilling Plan Generation Using Large Language Models

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

The oil and gas industry faces challenges in efficiently generating a comprehensive drilling plan, known as a 'design of service' (DoS) document, which requires significant time, effort, and resources due to its complexity and the potential for human errors, leading to delays and increased costs.

Innovation Solution

The implementation of a sequential data generator system utilizing large language models (LLMs) to automate the processing of raw data into formatted data, reducing human intervention and errors by generating a 'design of service' document, thereby streamlining the drilling plan creation process.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If traditional manual methods are used to generate design of service documents, then comprehensive human expertise and control are maintained, but significant time, effort, and resources are consumed leading to delays and increased costs

Engineering Contradiction:

Improvedrilling plan generation speedVSAvoidtime to prepare DoS document

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The patent replaces the manual mechanical process of drafting DoS documents with an automated system comprising a data processing module and a large language model. The system automatically retrieves well data from databases, processes it through structured prompts, and generates comprehensive drilling plans without manual intervention, thereby eliminating time losses associated with manual document preparation while maintaining comprehensive technical coverage.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system enables self-service generation of design of service documents by automatically retrieving necessary well data from databases, processing it through predefined technical frameworks, and producing complete DoS documents without requiring manual data collection or document assembly. The automated system serves itself by integrating data retrieval, processing, and document generation into a single autonomous workflow.

Inventive Principle:

Principle #25Self-service

2Reliability

If manual preparation of drilling plans is performed, then flexibility and expert judgment are maintained, but human errors increase leading to potential operational issues

Engineering Contradiction:

Improveaccuracy of drilling planVSAvoidhuman errors in DoS document

Core Design Contradiction:

ReliabilityVSObject-generated harmful factors

Solution Approach 1:

The patent eliminates human errors by replacing manual document preparation with an automated system that retrieves data directly from databases and processes it through structured large language model prompts. The system follows predefined technical frameworks and data processing rules, ensuring consistent and accurate generation of drilling plans without the variability and errors inherent in manual preparation.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system incorporates feedback mechanisms by retrieving actual well data from databases to validate and inform the generation of design of service documents. The large language model processes this real data through structured prompts, ensuring that generated plans are grounded in actual well characteristics and operational parameters, thereby improving reliability and reducing errors.

Inventive Principle:

Principle #23Feedback

3Device complexity

If comprehensive design of service documents are manually created, then complete drilling plans are produced, but the complexity and resource requirements increase significantly

Engineering Contradiction:

Improvecomplexity of DoS document preparationVSAvoidefficiency of drilling plan creation

Core Design Contradiction:

Device complexityVSProductivity

Solution Approach 1:

The patent segments the complex task of generating design of service documents into distinct functional modules: a data processing module that retrieves and organizes well data from databases, and a large language model that processes structured prompts to generate document content. This segmentation reduces the complexity of manual document preparation by dividing the workflow into automated, manageable components that can execute independently and efficiently.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system achieves multi-functionality by integrating data retrieval, data processing, document generation, and validation into a single automated platform. The large language model serves multiple purposes by processing various types of well data through different structured prompts to generate different sections of the drilling plan, thereby reducing the need for multiple separate tools and processes.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Data Source

PatentUS11966845B1Service document generation using large language models

Publication Date: 2024.04.23 HALLIBURTON ENERGY SERVICES INC

AI summary

A method for creating generated formatted data, that includes receiving, by a sequential data generator, raw data, where the raw data includes formation data at a drilling environment, processing the raw data to obtain generated recommendation data, where the generated recommendation data includes a proposed drilling location, and creating the generated formatted data, where the generated formatted data includes the generated recommendation data.