Identifying quantitative relationships between Key Performance Indicators in support of Physical Asset Management decision-making processes

[摘要] ENGLISH ABSTRACT: Physical Asset Management (PAM) is increasingly being acknowledged byindustry as an important contributor to the financial success of organisations, especially those who are dependent on their physical assets for organisationalvalue creation. Amongst the PAM improvement opportunities identified by researchers and organisations is the derivation of additional, meaningful and innovative information from Key Performance Indicators (KPIs) for improved PAM decision-making process.The Quantitative Relationships at the Performance Measurement System (QRPMS) methodology is an existing methodology which objectively identifiesand quantifies the relationships between a set of KPIs, and presents these relationships as additional information for PAM decision-making processes.QRPMS employs two mathematical techniques, Principal Components Analysis and Partial Least Squares regression, to identify and quantify inter-KPI relationships, respectively. The Guttman-Kaiser criteria (K1) is employed byQRPMS to determine the number of principal components (PCs) to retain for further assessment. However, the K1 criterion is found to be one of the least reliable and most inaccurate selection criteria available, with some publicationsusing it without reservation. Therefore, the K1 criterion severely compromises the reliability and mathematical accuracy of the results obtained from QRPMS.This study proposes an improved methodology for the objective identification and quantification of inter-KPI relationships, called the Quantitative Identification of Inter-Performance Measure Relationships (QIIPMR) methodology.A comprehensive literature study is conducted, investigating the realms of PAM, Performance Management (PM), Performance Management Systems (PMS) and performance measures. Existing frameworks and methodologieswhich aim to identify relationships between performance elements are investigated, and their flaws identified. The literature study concludes with an investigation of PCA, PLS and selection criteria. The proposed QIIPMRmethodology employs QRPMS as a foundational framework. Accurate and reliable alternatives to the K1 criterion are compared, and the most appropriate of these is incorporated into QIIPMR.A case study is conducted, comparing the results of QRPMS and QIIPMR using real-world KPI data from an open-pit, thermal coal mine in South Africa. The case study results substantiate the improvement made to QRPMSmethodology. This study concludes with recommendations for future research.

[发布机构] Stellenbosch University