3D Convolutional Neural Networks for Ore Grade Prediction

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Interpreting massive amounts of spatial geological information to identify ore bodies in underground mines is subjective, time-consuming, and requires expert knowledge specific to each mining project.

Innovation Solution

An ore content prediction system that receives structured geological data, trains a prediction model using multidimensional tensors derived from spatial information, and identifies relationships to predict the average grade of ore in a target region, suggesting revised input regions for improved accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If expert geologists manually interpret spatial geological information to identify ore bodies, then the interpretation can leverage expert knowledge specific to each mining project, but the process is subjective and time-consuming

Engineering Contradiction:

Improveinterpretation accuracyVSAvoidinterpretation time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent replaces the mechanical system of manual expert interpretation with an automated machine learning system. The system uses trained models to process spatial geological information and predict ore body locations, eliminating the need for time-consuming manual analysis while maintaining interpretation quality through algorithmic objectivity.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system enables self-service by allowing the machine learning model to automatically interpret geological data without requiring continuous expert intervention. The model trains on historical data and independently processes new spatial information to identify ore bodies, reducing dependency on expert time while preserving project-specific knowledge in the training data.

Inventive Principle:

Principle #25Self-service

2Loss of information

If massive amounts of spatial geological information are processed manually, then comprehensive analysis can be achieved, but the task requires many hours from individuals with expert knowledge

Engineering Contradiction:

Improveinformation completenessVSAvoidprocessing time

Core Design Contradiction:

Loss of informationVSLoss of time

Solution Approach 1:

The patent substitutes manual processing with automated computational methods. The machine learning system efficiently processes large volumes of spatial geological information including drill hole data, maps, and assay results, achieving comprehensive analysis without the time constraints of manual review by experts.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system segments the massive dataset into structured geological features that can be processed independently and systematically. By dividing the complex spatial information into manageable components such as lithology, structure, and mineralization data, the system achieves thorough analysis through organized computational processing rather than overwhelmed manual review.

Inventive Principle:

Principle #1Segmentation

3Adaptability or versatility

If subjective expert interpretation is used to identify ore bodies, then flexibility in decision-making is maintained, but consistency and objectivity are reduced

Engineering Contradiction:

Improvedecision flexibilityVSAvoidinterpretation consistency

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

The patent replaces subjective human judgment with objective algorithmic decision-making. The machine learning model applies consistent criteria across all data processing, eliminating variability introduced by different experts while maintaining adaptability through training on diverse, project-specific geological data that captures the nuances of different mining environments.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS12236329B2Estimate ore content based on spatial geological data through 3D convolutional neural networks

Publication Date: 2025.02.25 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

An ore content prediction system is provided. The system receives structured geological data that is derived based on spatial geological information that is associated with an input region. The received structured geological data includes a plurality of multidimensional tensors that are derived from spatial geological information of a plurality of sub-regions of the input region. The spatial geological information includes one or more types of data. The system trains a prediction model to produce a prediction output based on an average grade of an ore of a target mineral type at a target region by using the received structured geological data. The system identifies a relationship of the structured geological data to the prediction output and determines a revised input region based on the identified relationship.