Dealing with construction cost overruns using data mining

Construction Management and Economics (Impact Factor: 0.8). 06/2014; 32(7-8):682-694. DOI: 10.1080/01446193.2014.933854


One of the main aims of any construction client is to procure a project within the limits of a predefined budget. However, most construction projects routinely overrun their cost estimates. Existing theories on construction cost overrun suggest a number of causes ranging from technical difficulties, optimism bias, managerial incompetence and strategic misrepresentation. However, much of the budgetary decision-making process in the early stages of a project is carried out in an environment of high uncertainty with little available information for accurate estimation. Using non-parametric bootstrapping and ensemble modelling in artificial neural networks, final project cost-forecasting models were developed with 1,600 completed projects in this experimental research. This helped to extract information embedded in data on completed construction projects, in an attempt to address the problem of dearth of information in the early stages of a project. 92% of the 100 validation predictions were within ±10% of the actual final cost of the project whiles 77% were within ±5% of actual final cost. This indicates the model's ability to generalise satisfactorily when validated with new data. The models are being deployed within the operations of the industry partner involved in this research to help increase the reliability and accuracy of initial cost estimates.

Download full-text


Available from: Dominic D Ahiaga-Dagbui, Jun 24, 2014
548 Reads
  • [Show abstract] [Hide abstract]
    ABSTRACT: Purpose: Actual costs frequently deviate from the estimated costs in either favorable or adverse direction in construction projects. Conventional cost evaluation methods do not take the uncertainty and correlation effects into account. In this regard, a simulation-based cost risk analysis model, the Correlated Cost Risk Analysis Model, previously has been proposed to evaluate the uncertainty effect on construction costs in case of correlated costs and correlated risk-factors. The purpose of this paper is to introduce the detailed evaluation of the Cost Risk Analysis Model through scenario and sensitivity analyses. Design/methodology/approach: The evaluation process consists of three scenarios with three sensitivity analyses in each and 28 simulations in total. During applications, the model's important parameter called the mean proportion coefficient is modified and the user-dependent variables like the risk-factor influence degrees are changed to observe the response of the model to these modifications and to examine the indirect, two-sided and qualitative correlation capturing algorithm of the model. Monte Carlo Simulation is also applied on the same data to compare the results. Findings: The findings have shown that the Correlated Cost Risk Analysis Model is capable of capturing the correlation between the costs and between the risk-factors, and operates in accordance with the theoretical expectancies. Originality/value: Correlated Cost Risk Analysis Model can be preferred as a reliable and practical method by the professionals of the construction sector thanks to its detailed evaluation introduced in this paper.
    Engineering Construction & Architectural Management 07/2015; 22(4):403-423. DOI:10.1108/ECAM-09-2014-0121
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Construction projects routinely overrun their cost estimates. A plethora of studies have thus been dedicated to investigating the root causes, sizes, distribution and nature of overruns. The causes range from a poor understanding of the impact of systemicity and complexity of projects, unrealistic cost targets and misguided trade-offs between project scope, time and cost to suspicions of foul play and even corruption. In spite of the vast attention dedicated to the problem of cost overrun, there has been limited evidence to support the claim that the size or occurrence of cost overruns is reducing in practice. A review of the literature reveals that it may not be an exaggeration to claim that the bulk of our current cost overrun research may largely be inadequate and deficient to deal with the complexity posed by construction projects. This paper provides a critique of current cost overrun research and suggests that the adoption of systems thinking is required to better understand the nature of cost overruns. We explore some of the embedded methodological weaknesses in the approaches adopted in a majority of cost overrun research, particularly the lack of systems thinking and demonstrable causality. We reach the following conclusion - cost overrun research has largely stagnated in the refinement and advancement of the knowledge area. It has largely been superficial and replicative. A significant paradigm and methodological shift may be required to address this perennial and complex problem faced in construction project delivery
    Procs 31st Annual Association of Researchers in Construction Management Conference, Lincoln, UK; 09/2015
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Megaprojects present significant economic risks to their financiers and sponsors. Factors such as inflation, cash flow issues, material and energy price hikes and change in government policies can cause such capital intensive projects to overrun planned budget and schedule allocations. Where the project is a commercial asset, delayed completion time and cost overruns usually have significant impact on the profitability of the project as well as the estimated returns on investment over the operational phase of the project. Understanding the dynamics of specific risks can thus be very crucial in designing containment measures to deal with their likely impact on the project. Using a case-study of the Edinburgh Trams project in Scotland, the dynamics of identified economic risks in transportation megaprojects is presented. Through the combination of interviews, questionnaires and non-participant observation, different economic risk factors were first identified. The identified factors were then prioritised using Analytical Network Process (ANP) to establish the most salient economic variables on the Tram project. Some of these factors include material and energy price increases as a result of the 2008 recession, as well as inflation and changes in government funding policies. The selected factors from the ANP were then modelled within a System Dynamics (SD) framework to appraise their measured economic impact on the project to gain a fuller understanding of the interrelationships between the variables in the system. The mean impact of economic risks on Edinburgh Trams was estimated to be about 22%.
    Procs 31st Annual Association of Researchers in Construction Management Conference,, Lincoln, UK; 09/2015