Development of a Markov model for production performance optimisation. Application for semi-automatic and manual LHD machines in underground mines

Abstract

This paper compares three ways to operate a load haul dump (LHD) machine, manual operation, automatic operation (fleet operation) and semi-automatic operation, to find the best operating mode. In a fault tree analysis, different failures are classified and analysed, but the way to recover from certain states is not accounted for, which is something a Markov model can handle. The paper is based on the analysis of real data from an underground mine. A Markov model has been built for mining application and it is shown that a semi-automatic LHD has the highest probability of being in a productive state since it has the advantage of changing operating modes (manual and automatic) depending on the need and situation. Hence, the semi-automatic LHD is the best choice from an operational point of view. The paper fills a gap in the literature on manual vs. automatically operated LHDs by providing a new way of evaluating the operating mode of LHDs using Markov modelling, while considering the operating environment.