This paper describes an innovative method to calculate the quality of a stockpile in real-time to improve the efficiency in stockpile blending operations. A stockpile is modelled as a grid of voxels (octants), which are used as geometric elements for both quality calculations and result storage. The quality calculation is based on the real geometric shape of the stockpile which is measured by a laser scanner. Furthermore, a BWR (Bucket Wheel Reclaimer), which is treated as a mobile robot arm, is incorporated into this model to predicate the quality of the reclaimed material. Thus, this 3D quality embedded model can be used to calculate, plan, control and predict the quality level of iron ore being stacked and/or reclaimed with a great degree of accuracy. This model provides an opportunity to monitor and track quality data throughout the blending operations. It will lead to proactive and continuous blending activities that fundamentally improve quality control and increase efficiency in stockpile management. (C) 2015 Elsevier B.V. All rights reserved.