According to the abstract, "a decision support system for underground mining method selection (UMMSDSS) has been designed and developed in order to take into account the whole related problem criteria, research the entire effects of different scenarios of all available criteria and carry out sensitive analysis when needed." The description notes that the system uses multi‑criteria decision‑making (MCDM) techniques and is aimed at "the evaluation and selection of the optimal mining method" for metallic mineral deposits under varying conditions.

The paper states: "The primary objective of mining methods selection (MMS) is to select method(s) that maximize profitability, operational efficiency, safety, and environmental sustainability." It proposes "an AI-based approach for developing a recommendation system for MMS" and explains that "the developed system uses artificial intelligence techniques to recommend the most appropriate mining method(s) for a given ore deposit based on its characteristics." The authors indicate that by learning from historical cases, the AI system "can predict the optimal mining method" under varying geological and operational conditions.

This peer-reviewed study describes "a fuzzy techno-financial methodology" for mining method selection and states: "The objective of mining method selection is to choose the optimal method for a deposit considering both technical and economic factors." The method combines fuzzy logic with techno‑economic analysis to "handle uncertainties in ore body characteristics" and to evaluate alternative methods. While not a neural-network system, it shows the use of computational intelligence and fuzzy decision tools to "determine the optimal mining method" for a specific deposit case study.

This review paper summarizes: "In this paper, the application of deep learning in the mining and processing of ores is reviewed. Deep learning is strongly impacting the development of sensor systems, particularly computer vision systems used in mining and mineral processing automation." It notes that, beyond perception tasks, "deep learning is also being considered in the automation of decision support systems," and concludes that there is "significant scope for the application of deep learning to improve operations," although access to industrial data is a key bottleneck.

The paper notes: "There is no single appropriate mining method for a deposit. Usually two or more feasible methods are possible." It introduces the Analytic Hierarchy Process (AHP) as a multi-attribute decision-making technique and explains that it can be used to select a mining method by systematically ranking alternatives. The authors state that, unlike traditional approaches, "AHP makes it possible to select the best method in a more scientific manner" and that the process "provides a prioritized rank order indicating the overall degree of preference for each decision alternative" for a given deposit. Although AHP is not machine learning, it is a formal algorithmic decision method used to select an optimal mining method based on multiple criteria.

The paper reviews and applies decision-making tools, including artificial intelligence-based models, to underground mining method selection. Using a case study, the authors compare the performance of classical scoring methods with expert systems and fuzzy logic approaches. They conclude that AI-based decision tools better handle the complexity and uncertainty of deposit parameters and can more consistently identify the most suitable mining method.

The paper discusses open-pit production scheduling and states that "an artificial intelligence (AI) based methodology is proposed" to obtain operative pushbacks in open-pit mines. It explains that the approach uses "Genetic Algorithms (GAs) and a clustering algorithm (k-means)" to respect operational and design constraints while maximizing net present value. Although the focus is on phase design and scheduling rather than method selection per se, it is a concrete demonstration that AI techniques (GA and k-means) can be used in mine design optimization problems related to how a deposit is mined.

This conference paper explains that "selection of an underground mining method is a key decision in mine planning" and that the decision depends on many geological and technical parameters. The authors propose "the use of artificial neural networks (ANNs)" to support this decision, training the network on data from existing mines. They report that the ANN model "was able to predict the suitable underground mining method" for new cases based on orebody characteristics, illustrating direct use of AI (neural networks) to select an appropriate method.

The CORDIS article describes an EU-funded project that develops new machine learning methods to analyse Earth observation data for mineral exploration. According to the project coordinator, these weakly and self-supervised tools can 'identify mineral deposits and monitor environmental impacts' and 'enhancing mineral exploration activities.' The article explains that such tools are designed to support mining operations by providing better information about mineral deposits, which feeds into downstream decisions such as mine design and method selection.

The article discusses how artificial intelligence is being embedded in "decision support tools for climate risk, mineral exploration and mining". It explains that AI-driven software can help operators evaluate multiple mine design options and scenarios, and it notes emerging tools that allow engineers to test different mine plans and configurations digitally in order to choose the most efficient and economic approach for a given orebody.

This industry article states that "industrial AI is transforming mining" and describes how AI-powered systems are used to "predict equipment failures, optimize crushing and grinding energy use, stabilize throughput despite ore variability, and enhance safety in hazardous environments." It notes that AI is being applied "across the value chain, from exploration targeting and autonomous haulage to AI for mineral processing," highlighting that AI methods are already used for complex optimization and decision tasks in mining operations, even if it does not specifically discuss mining method selection.

This industry-focused overview explains that "Deep learning uses artificial neural networks to learn from data. When trained on large, high-quality datasets, it achieves high accuracy, sometimes exceeding human-level performance in specific tasks." Among listed use cases are several in heavy industry and resource extraction, and the article notes that deep learning is increasingly used "to optimize complex operational decisions" in domains where many variables interact.

Since the late 1990s and early 2000s, mining engineering literature has documented the use of expert systems, fuzzy logic, neural networks and hybrid AI models to assist in mining method selection. These systems typically encode empirical rules relating deposit geometry, rock mass quality and economic factors to feasible methods, and then use AI algorithms to rank or classify alternatives. While they are decision-support tools rather than fully automated optimizers, they are explicitly developed to help select an 'optimal' or 'most suitable' mining method for a given mineral deposit.

In this 2025 industry talk on the "state of AI in the mining industry", the speaker describes how combining digital twins with large language models can create a "decision support layer" for mine planning. He explains that by integrating sensor data, geological models and design information, such AI systems allow engineers to "test more scenarios faster" for drilling, blasting and mine design choices, supporting optimized mining strategies for particular operations.