Analysis of Existing Resource Distribution Metrics and Suggestion of New Metrics

Abstract

Resource leveling aims to reduce peak resource demand and fluctuations in resource usage histograms throughout a construction period. The objective function optimum can be obtained in the digitalized construction management era. However, the correct quantification of uneven resource distribution is also essential because an improper resource distribution metric may even worsen the resource distribution. The behaviors of existing resource distribution metrics were not examined thoroughly in the literature. This study aims to improve existing resource distribution metrics and analyze the objective functions to identify their positive and negative aspects. Eight existing resource leveling metrics were analyzed by considering their responses to the peak resource demand and fluctuations in daily resource demand. The analysis uncovered that the existing metrics were insensitive to either resource fluctuations or peak resource demand. The time insensitiveness of the minimum moment metric was eliminated by merging it with the sum of the square metric. In addition, a fractional order resource distribution metric was proposed to dissuasively penalize peak resource demand. A comparison of resource distribution metrics for two case study problems revealed that the proposed three resource distribution metrics presented better resource distribution histograms with less fluctuations and peak resource demand. This study revealed the behavior and disadvantages of existing resource distribution metrics such that the schedulers can foresee the probable output of the metrics, which may assist them in implementing a proper metric. Moreover, the metrics developed in this study provided better resource histograms, which was the theoretical contribution of this study.