The key skills for a competent geologist are observation, documentation, analytical thought and the ability to communicate. These attributes have served the geological world for the past two centuries since the field of geology was separated out from natural philosophy. The understanding of the structural setting, lithologies, alteration and 3D distribution of the target mineral species are essential precursors to resource evaluation and potential economic extraction. The advent of artificial intelligence and machine learning has matured to the point where observation by dedicated sensors and analysis by customised AI algorithms can replace manual processes and deliver credible results. In combination with analytical data such as assays, various sensors now collect physical properties of rocks which are used as proxies for mineral identification and rock mass classification. These sensors include X-ray and infra-red spectra, magnetic susceptibility, gravity/density, EH/pH measurements, conductivity, colourimetric readings, porosity, fracturing and fracture intensity. Analysis using machine learning techniques trained using known data sets can determine rock type classification, rock density, quantification of grade/quality attributes, mineral alteration, weathering and geotechnical attributes such as rock quality. This type of data relies on calibration and quality control of a variety of different types of electronic hardware. The resulting data is non-additive, has error ranges and will be suitable for consumption by fast AI methods providing probabilistic analysis. A second stage could include geometallurgical constraints to provide an ‘extractability index’, which is of more interest to miners than pure geotyping. Using resultant interpreted geological criteria, 3D domain models with spatial attributes will be built.