Data analysis and prediction models in oil sands mining

Proper operation and maintenance of equipment are essential to effective mining operations in oil sands industry relying heavily on truck and shovel methods. The operational goal is to minimize truck damage while maximizing production. To effectively implement this understanding, continuous monitoring and predictive modeling of truck shovel parameters is needed. Of paramount importance is to predict heavy truck drivers' exposure to vibration during regular tar sand hauling operation.; This research focuses on analysis and modeling of experimental datasets collected from tar sand heavy hauler operations in summer conditions. The stress-related accelerations were measured by tri-axial accelerometers mounted on the seat pan of the truck driver. Several predictive models were developed to investigate the effect of important input parameters of a hauling truck (suspension cylinders pressure, ground speed, payload weight, rack and pitch) on these accelerations magnitudes. Best performing model was found experimentally, able to predict acceleration changes with reasonable accuracy.