The rising demands for mined products lead to the extraction of materials in geologically complex regions. This calls for mining process changes and interventions driven by technology and advanced data analytics. The dynamic development of state-of-the-art sensor technologies and their potential use in mining is projected to significantly reduce costs in the industry. However, despite rapid advances in sensor technologies, there is still a demand for novel data analytical approaches to enable accurate characterisation of material along the mining value chain, as advanced data analytics is key to gain knowledge from the complex sensor-derived data. Therefore, sensor technology, coupled with advanced data analytics is crucial for the rapid and accurate characterisation of material in mining operations. Access to rapid and accurate data on the key geological attributes (e.g., mineralogy and geochemistry) along the mining value chain has significant implications for the production process efficiency in commercial mines. Such data would greatly assist the improvement of deposit models, optimise ore processing, specify product quality and improve operational decision-making. Sensor technologies operate over a specific range of the electromagnetic spectrum and provide information on certain aspects of material properties that are of potential interest for mining extraction. However, a single sensor might not provide a sufficiently comprehensive description of a material’s composition. This introduces uncertainty into both resource estimation and requirements definition for mineral processing. Thus, it is necessary to utilise strategic sensor combinations to improve accuracy, minimise uncertainty, and enhance specific insights of material compositions. Combinations of sensors can be implemented using a data fusion approach. The fusion of sensed data can be realised at different levels: low-, mid-, and high-level, when the integration occurs at the data level, features level and decision level, respectively. This research aims to develop methods for the characterisation of raw materials using multiple sensor technologies and sensor combinations concept (data fusion at different levels), that can be potentially applicable to mining operations. The study involved the multispectral and hyperspectral imaging techniques, such as red-green-blue (RGB) imaging, visible and near-infrared (VNIR) and short-wave infrared (SWIR) hyperspectral imaging, and point spectroscopic techniques, such as mid-wave infrared (MWIR), long-wave infrared (LWIR) and Raman spectroscopy to acquire spectral information over a wider range of the electromagnetic spectrum. First, an investigation was conducted on the usability of the individual sensor technologies coupled with data analytics for the characterisation of a polymetallic sulphide deposit at different levels. The different levels of material characterisation aimed to allow mineral mapping, ore–waste discrimination, fragmentation analysis, and semi-quantitative analysis of elements and minerals. The positive outcomes of the use of the individual techniques led to the development of a data fusion framework that enables data integration (including multi-scale and multi-resolution data) at different levels (e.g., low-level and mid-level). The developed data fusion concept was implemented and validated using different test scenarios...