OEE improvement of mining shovels by survival analysis and linear optimisation as per sustainable development goals

Abstract

ABSTRACT Sustainability is acknowledged as an emerging megatrend in business that significantly affects companies’ survival and competitiveness in the market-place. Environmental, global workforce and complex supply chain networks have created pressures to have a clear vision and improve sustainability due to geopolitical dimensions. As per United Nations (UN), Sustainable Development Goals (SDGs), responsible production and consumption (SDG 12), and industry, innovation and infrastructure (SDG 9), and attain higher economic scales of productivity (SDG 8) oblige to drive productivity improvement. The mining maintenance costs constitute around 30% to 40% of the direct mining costs in mining due to diverse operating conditions. First, this article aims to develop the Cox regression Machine Learning (ML) model to derive shovels’ Remaining Useful Life (RUL). Second, formulate and model the maintenance schedule optimisation of mining equipment. Third, test and validate the cost optimisation model by deploying Decision Optimisation (DO) ILOG CPLEX to combine maintenance schedules of Preventive Maintenance (PM) and Predictive Maintenance (PdM). Finally, the data-driven actions demonstrate operating cost reduction through the metrics of Overall Equipment Effectiveness (OEE), Overall Throughput Effectiveness (OTE) and Impact Factor (IF) computation. Further, this article demonstrates the benefits of IF improvements through a case study. The combined optimised maintenance reduced shovels’ maintenance by 2.27 hours per combined schedule, which led to the potential ‘OEE’: improvement between 2.7% and 7.2% of different shovels and which is a measure of equipment productivity. The computed IF improvement for mining shovels is 49% and is aligned as per the SDG of responsible production.