<p>Current methods for mining low-grade deposits are expensive and resource-intensive, requiring vast amounts of energy, water, and land use. These methods often involve hazardous chemical processes as well as the creation of large tailings ponds. Pre-concentration is the removal of gangue prior to processing and offers a more sustainable mining alternative by reducing the volume of rock subjected to downstream processing, offering both economic and environmental benefits. Minerals exhibit diverse properties (e.g., density, gamma-ray emissions, optical, thermal, magnetic, and structural) that can be exploited using various pre-concentration technologies. Advances in smart sensors, Internet of Things (IoT), Artificial Intelligence (AI), and wireless communication have enhanced industrial sorting, enabling real-time monitoring of mineral characteristics. Selecting the optimal sensor and sorting algorithm depends on aligning mineralogical differences with sensor capabilities. Understanding sensor measurement methods, resolution, and deployment is essential for effective application. To date, the application of automated mineralogy images and centimeter (meso) -scale analysis for pre-concentration and sorting has not been fully investigated. The goal of this work is to demonstrate the potential of meso-scale analysis by presenting the rationale for sensor selection in three case studies, each with a different target commodity: an orogenic gold deposit, a volcanic massive sulfide (VMS) deposit, and a rare earth element (REE) deposit. These case studies demonstrate that appropriate sensor selection can remove 50–98% of gangue and highlight the importance of considering mineral properties, meso-scale homogeneity, sensor response time, and deposit geology when choosing sensors.</p>