KPI framework for maintenance management through eMaintenance : Development, implementation, assessment, and optimization

Abstract: Performance measurement is critical if any organization wants to thrive. The motivation for the thesis originated from the project “Key Performance Indicators (KPI) for control and management of maintenance process through eMaintenance (in Swedish: Nyckeltal för styrning och uppföljning av underhållsverksamhet m h a eUnderhåll)”, initiated and financed by a mining company in Sweden. The main purpose of this project is to propose an integrated KPI system for the mining company’s maintenance process through eMaintenance, including development, implementation, assessment, and optimization.There are gaps in the research, however, resulting in the following challenges. First, no KPI framework considers both technical and soft KPIs, so developing a system is problematic. Second, few studies have focused on implementing KPI measurement through eMaintenance. Third, there are gaps in KPI assessment. In assessing system availability, for example, the current analytical (e.g., Markov/semi-Markov) or simulation approaches (e.g., Monte Carlo simulation-based) cannot handle complicated state changes or are computationally expensive. In addition, few researchers have revealed the connections between technical and soft KPIs.  For those soft KPIs for which the distribution of data collected from eMaintenance systems (e.g., work orders) is not easily determined, studies are insufficient. Fourth, the current continuous improvement process for the KPIs is very time-consuming. In short, there is a need for a new approach.The thesis develops an integrated KPI framework consisting of technical KPIs (linked to machines) and soft KPIs (linked to maintenance workflow) to control and monitor the entire maintenance process to achieve the overall goals of the organization.  The proposed KPI framework makes use of four hierarchical levels and has 134 KPIs divided into technical and soft KPIs as follows: asset operation management has 23 technical KPIs, maintenance process management has 85 soft KPIs and maintenance resources management has 26 soft KPIs.The thesis discusses the proposed KPI framework; it lists the KPIs and provides timelines, definitions and general formulas for each specified KPI. Results will be used by the mining company to guide the implementation of the proposed KPIs in an eMaintenance environment.To suggest novel approaches to KPI assessment, the thesis takes system availability in the operational stage as an example.  It proposes parametric Bayesian approaches to assess system availability. With these approaches, Mean Time to Failure (MTTF) and Mean Time to Repair (MTTR) can be treated as distributions instead of being “averaged” by point estimation. This better reflects reality.  Markov Chain Monte Carlo (MCMC) approach is adopted to take advantage of both analytical and simulation methods. Because of MCMC’s high dimensional numerical integral calculation, the selection of prior information and descriptions of reliability/maintainability can be more flexible and realistic. The limitations of data sample size can also be compensated for. In the case studies, Time to Failure (TTF) and Time to Repair (TTR) are determined using a Bayesian Weibull model and a Bayesian lognormal model, respectively. The proposed approach can integrate analytical and simulation methods for system availability assessment and could be applied to other technical problems in asset management (e.g., other industries, other systems). By comparing the results with and without considering the threshold for censoring data, the research shows the connection between technical and soft KPIs, and suggests the threshold can be used as a monitoring line for continuous improvement in the mining company. For those soft KPIs for which the distribution of data collected from the eMaintenance system (e.g., work orders) is not easily determined, other approaches, such as time series analysis (if the data are “fast moving”), the Croston method (if the data are “intermittent”), or the bootstrap method (if the data are “slow moving”) could be applied. To ensure the KPI framework can be improved continuously, the thesis performs a comparison study to find the gaps between current and proposed KPIs in the mining company. It adapts a roadmap from the railway industry to show how optimization can be promoted by reviewing and improving the KPI framework.Results from this study will be applied to the company and guide its development, implementation and assessment of the KPIs through eMaintenance with continuous improvement. The proposed approaches could also be applied to other technical problems in asset management (e.g., other industries, other system).