Artificial Intelligence, Geometallurgy and Standards for Communication between Humans and Machines
Clinton Smyth, Minerva Intelligence Inc; Samuel Cantor, Minerva Intelligence
Artificial Intelligence (AI) is a broad field which encompasses machine learning, knowledge representation, robotics, probability and a host of other disciplines and sub-disciplines. Fundamental to the successful application of AI in the mining industry is the establishment of clear communication between human beings using AI and the computers which deploy or generate that AI – both humans and computers being considered as “agents” in the field of AI. Humans need to clearly describe what they provide as input to AI systems, and AI systems need to be able to explain in human-intelligible terms what they produce and how they produce it. For clarity of communication between agents it is essential that the terms used in that communication are understood by all agents, both human and computer, to have the same meaning. This clarity of communication is achieved by using standardised terminologies, many of which are currently under development in fields relevant to geometallurgy, such as the taxonomies of rocks, minerals, processing chemicals and pollutants. In the European Union, compliance with some of these standards has already been incorporated, under the EU INSPIRE Directive, into legislation governing reporting obligations. This presentation describes a standards-dependent geometallurgical AI application developed for finding, for a newly-discovered mineral deposit, the metallurgical treatments historically implemented for mineral deposits of similar lithology, mineralogy and geochemistry, based on geometallurgical considerations. It also shows how the same standard terminologies necessary for geometallurgical AI applications are applicable to both AI-based minerals exploration and to AI-based applications in environmental protection. The presentation concludes by observing that this linking of minerals exploration, extractive metallurgy and environmental protection via the terminologies that they use constitutes “best practice” in the modern world of knowledge economies, and is achievable by the adoption of semantics-aware software systems already used by many of the world’s larger corporations.
Artificial Intelligence, metallurgy, geometallurgy , Big Data, IoT, Semantics, Standards, Machine Learning, Graph Databases, Ontology, Taxonomy, Exploration, Mining, Natural Hazards, Geospatial