The focus in the iron ore industry has shifted from increasing production volumes with less regard for cost, to efficient production of quality tonnes. BHP Billiton iron ore geoscience and exploration has embedded infrared spectroscopic sensing of drill hole samples to help meet this challenge.Infrared spectroscopy is delivering mineral predictions that are used for (automated) geology logging and modelling. The spectra are processed together with other available data sets to predict geometallurgical and geotechnical variables that include lump per cent and uniaxial compressive strength.Trials are ongoing to test the benefit of infrared spectroscopy during mining, including the analysis of grade control samples and remote sensing of mine faces. Results show great benefits in the fields of short-term geology modelling, better understanding of material (handling) properties, identification of geotechnical risk areas and high resolution ore-waste delineation.These trials are bringing BHP Billiton iron ore closer to its ultimate goal of having infrared spectroscopic data collected across the mining value chain. Once such data are available everywhere, regression models can be trained at one point (eg iron grade model trained on exploration samples) and be applied elsewhere (eg iron grade predicted on mine faces). Such predictive analytics will enable not only high precision mining, but also provide maximum time to tailor plant design to the ore in the ground, based on learnings from ores processed in other plants. This will improve planning outputs that will eventually increase mining efficiency, processing capability and product quality.CITATION:Haest, M, Mittrup, D and Edwards, L, 2017. Infrared spectroscopy in BHP Billiton iron ore with a focus on in-mine infrared sensing, in Proceedings Iron Ore 2017, pp 549-556 (The Australasian Institute of Mining and Metallurgy: Melbourne).