Predicting construction plant maintenance expenditure

D. J. Edwards, G. D. Holt

Research output: Contribution to specialist publicationArticle

6 Citations (Scopus)

Abstract

Utilization of construction plant and equipment forms an essential part of construction contractors' efforts to augment productivity and reduce costs. However, efficient and effective management of plant and equipment is requisite to maximizing its potential benefits. A problem in this respect is now knowing when a maintenance event might occur, or its extent in terms of cost and time. This paper presents a stochastic mathematical modelling methodology (using random numbers) to predict the probable cost of 'the next' maintenance event for tracked hydraulic excavators; where a maintenance event includes both breakdown and scheduled maintenance. Information on a random sample of 33 360° tracked hydraulic excavators were obtained from contractors operating within the opencast mining industry. From these, 9473 maintenance event 'cost' observations were recorded and modelled. Validation of the model is achieved using a random 'hold out' sample of 50 maintenance cost observations taken from nine machines. Analyses reveal that overall model predictive performance is robust, having a mean percentage error (MPE) of 4.46 and a mean absolute percentage error (MAPE) of 23.63. Pearsons correlation coefficient (r) and a paired t-test are conducted to determine the accuracy and consistency of model predictions respectively. With an r of 0.76 and no significant difference being found between the mean of predicted/actual values, the model shows both accurate and consistent cost predictions. The random number technique shows potential for improved maintenance practice by providing a practical methodology for planning, scheduling and controlling future plant resource requirements. The paper concludes with direction for future research which includes: (1) the application of this research to plant items working in other operational environments (e.g. civil engineering and construction); and (2) prediction of the next breakdown event.

Original languageEnglish
Pages417-427
Number of pages11
Volume29
No.6
Specialist publicationBuilding Research and Information
DOIs
Publication statusPublished - Nov 2001
Externally publishedYes

Keywords

  • Construction plant
  • Financial management
  • Linear interpolation
  • Maintenance
  • Management
  • Prediction
  • Stochastic random numbers
  • UK

ASJC Scopus subject areas

  • Civil and Structural Engineering
  • Building and Construction

Fingerprint

Dive into the research topics of 'Predicting construction plant maintenance expenditure'. Together they form a unique fingerprint.

Cite this