Content
Pepole
CV
Amirkabir University Researchers Develop an Intelligent Copper Exploration Model Using AI and Geological Data
| Last update: 2026/01/27 |
A graduate of Amirkabir University of Technology, under the supervision of the university's professors, has presented a novel AI-based framework for the accurate prediction of porphyry copper mineralization potential zones. He stated that this model addresses a significant gap in previous studies by enabling the integration of geological data and advanced algorithms, which can substantially reduce the cost and risk of mineral exploration operations.
According to the Public Relations Department of Amirkabir University of Technology, Dr. Yusuf Bahrami, supervised by Dr. Hossein Hasani and advised by Dr. Abbas Maqsoudi, faculty members of the university's Mining Engineering Faculty, implemented a project titled "Optimization of Integrated Predictive Models for Porphyry Copper Mineralization Potential Zones Based on the Application of Novel Algorithms."
Bahrami cited the existence of a clear research and operational gap in the systematic exploration of porphyry copper deposits in Iran, especially in the strategic and high-potential Paryz-Chahargonbad region of Kerman, as the primary reason for choosing this research.
He identified this gap across several main axes, one being the dispersion and lack of coherence in previous studies. He stated: "Although scattered studies had been conducted on this region in the past, these researches were often sectional, limited to a specific domain (e.g., only geochemistry or only remote sensing), and lacked an integrated and comprehensive approach. It was as if each piece of a puzzle was examined separately, but a complete and reliable picture of the entire region was never provided. This dispersion led to the inaccurate identification of promising areas and increased investment risk in exploration."
This Amirkabir University of Technology graduate considered the non-utilization of novel approaches, especially advanced AI-based approaches, as one of the issues in this field, adding: "Previous studies mainly relied on traditional and classical methods and did not significantly benefit from the power of advanced artificial intelligence algorithms, machine learning, and deep learning for analyzing complex geological data. However, the geological structures controlling mineralization are highly complex and nonlinear, and processing them requires powerful computational tools."
Bahrami, stating that the goal was to use these approaches to reveal hidden patterns in data that remain unseen by traditional methods, mentioned the urgent need to reduce the cost and risk of exploration operations as another gap in this field. He added: "The mineral exploration process is very costly and risky. The higher the accuracy of predictive models, the more precisely and accurately target areas for subsequent studies (such as field controls, sampling, and drilling) are determined, thereby preventing unnecessary costs. Therefore, the main objective of this research was to develop an intelligent integrated model that, by integrating all available data and processing it with novel and creative algorithms, provides a highly reliable roadmap for exploration teams so they can focus their resources and time on the most promising areas."
He described the goal of this research as responding to a scientific and practical need in the mining industry: how to find hidden treasures of copper deep within the earth with unprecedented accuracy before drilling and optimize the costly and time-consuming exploration process. He added: "The aim of this thesis was to create an integrated framework in which we first performed maximum optimization in each data section (from extracting hydrothermal alterations using advanced methods and deep networks, identifying and refining faults using satellite images, preparing detailed lithology maps, to separating geochemical anomalies using complex nonlinear methods)."
Bahrami continued: "Then, in the final stage of integrating evidence layers from different data sources, using adva
nced methods, including graph deep learning methods, we achieve maximum optimization. This two-stage strategy (i.e., layer optimization followed by integration optimization) has led to a tangible reduction of uncertainty at each step and consequently enhanced the accuracy and reliability of the resulting potential models."
The Amirkabir University of Technology graduate cited advancing the frontiers of knowledge and methodological innovation as achievements of this research, stating: "This research has advanced the frontiers of knowledge in the field of Mineral Potential Modeling (MPM) by introducing and applying a completely innovative framework."
According to him, the multi-stage and hybrid combination of methods includes: fusion of various satellite images with optimal approaches, extraction of hydrothermal alterations from satellite images using precise sub-pixel methods, preparation of high-detail geological maps by designing and developing advanced hybrid deep architectures such as CNN-GoogleNet-HHO, designing and developing an advanced processing chain such as Autoencoder-UMAP-GAN-Bisecting K-means for separating geochemical anomalies, and utilizing advanced methods such as Graph Deep Learning Networks (e.g., GAT) for the final integration and modeling of copper potential.
Bahrami stated that this achievement directly impacts the reduction of major exploration sector costs, and listed reducing exploration-related risks and costs, increasing the effectiveness of exploration investment, and high commercialization potential among the achievements of this project.
He emphasized: "Iran, with its vast mineral reserves, has always faced the challenge of systematic and low-cost exploration of these reserves. This model introduces a new standard method that can also be generalized as a pattern for other mineral regions of Iran (such as iron, gold, and lead-zinc belts). This research contributes to increasing the discovery rate of new mineral deposits in the country, which is the foundation for industrial development and the creation of national wealth."
The project executor considered the most important feature of this research to be its high reliability in identifying potential areas, noting: "In the past, decisions to continue exploration and drilling operations were based on qualitative and semi-quantitative interpretations associated with high uncertainty. This model provides a strong quantitative and statistical foundation for mining managers' decisions. Also, by automating a large part of data processing, the speed of analysis and preparation of mineral potential maps increases significantly."
He described the results of this research as an interdisciplinary achievement and said:"The main users of this project include: the mining and mineral exploration sector, engineering consulting companies, the Geological Survey and Mineral Exploration Organization of Iran, National Iranian Copper Industries Company, the Ministry of Industry, Mine and Trade, IMIDRO, Iranian Mines and Mining Industries Development and Renovation Organization, Mineral Production and Supply Company, space and remote sensing organizations, electronics and cable manufacturing industries, and universities and research institutes."
He added: "This project benefits from advantages such as providing an integrated and unique solution, modular and expandable design, reduced exploration cost and time, exceptional accuracy and expandability, and the possibility of rapid adaptation of the framework for other mineral regions and elements like gold, lead and zinc, iron, and even strategic elements like lithium and rare earth elements."
The Amirkabir University of Technology graduate cited advancing the frontiers of knowledge and methodological innovation as achievements of this research, stating: "This research has advanced the frontiers of knowledge in the field of Mineral Potential Modeling (MPM) by introducing and applying a completely innovative framework."
According to him, the multi-stage and hybrid combination of methods includes: fusion of various satellite images with optimal approaches, extraction of hydrothermal alterations from satellite images using precise sub-pixel methods, preparation of high-detail geological maps by designing and developing advanced hybrid deep architectures such as CNN-GoogleNet-HHO, designing and developing an advanced processing chain such as Autoencoder-UMAP-GAN-Bisecting K-means for separating geochemical anomalies, and utilizing advanced methods such as Graph Deep Learning Networks (e.g., GAT) for the final integration and modeling of copper potential.
Bahrami stated that this achievement directly impacts the reduction of major exploration sector costs, and listed reducing exploration-related risks and costs, increasing the effectiveness of exploration investment, and high commercialization potential among the achievements of this project.
He emphasized: "Iran, with its vast mineral reserves, has always faced the challenge of systematic and low-cost exploration of these reserves. This model introduces a new standard method that can also be generalized as a pattern for other mineral regions of Iran (such as iron, gold, and lead-zinc belts). This research contributes to increasing the discovery rate of new mineral deposits in the country, which is the foundation for industrial development and the creation of national wealth."
The project executor considered the most important feature of this research to be its high reliability in identifying potential areas, noting: "In the past, decisions to continue exploration and drilling operations were based on qualitative and semi-quantitative interpretations associated with high uncertainty. This model provides a strong quantitative and statistical foundation for mining managers' decisions. Also, by automating a large part of data processing, the speed of analysis and preparation of mineral potential maps increases significantly."
He described the results of this research as an interdisciplinary achievement and said:"The main users of this project include: the mining and mineral exploration sector, engineering consulting companies, the Geological Survey and Mineral Exploration Organization of Iran, National Iranian Copper Industries Company, the Ministry of Industry, Mine and Trade, IMIDRO, Iranian Mines and Mining Industries Development and Renovation Organization, Mineral Production and Supply Company, space and remote sensing organizations, electronics and cable manufacturing industries, and universities and research institutes."
He added: "This project benefits from advantages such as providing an integrated and unique solution, modular and expandable design, reduced exploration cost and time, exceptional accuracy and expandability, and the possibility of rapid adaptation of the framework for other mineral regions and elements like gold, lead and zinc, iron, and even strategic elements like lithium and rare earth elements."
.jpg)
.jpg)
Related
View: 244 Time(s) |
Print: 31 Time(s) |
Email: 0 Time(s) |
0 Comment(s)