DEVELOPMENT OF A BIG DATA AND PREDICTIVE ANALYTICS CAPABILITY ASSESSMENT TOOL (BDPA-CAT) FOR THE CONSTRUCTION INDUSTRY

Abstract

The concept of big data can be concisely summarized in four themes - data characteristics, technologies, analytical methods and impact to the society. Analytics comprises a series of statistical methods and categories. Predictive analytics is associated with a higher level of complexity and value to the more conventional descriptive and diagnostic analytics. Despite of the hype for big data and predictive analytics (BDPA) in the recent decade, development of BDPA in the construction industry is significantly falling behind other industries such as supply chain, healthcare, insurance and e-commerce. This study aims to clearly define BDPA in the construction industry by exploring its key processes, techniques and technologies. It also assessed the status quo of BDPA implementation in the local construction industry through a pilot survey and a main survey with 42 experienced professionals from the industry. Frequency analysis, Shapiro-Wilk test, one-sample Wilcoxon signed rank test, Kruskal-Wallis one-way analysis, Mann Whitney U test and mean rank analysis had been performed on the survey responses to draw deeper insights. From extensive literature review, interviews and survey analysis results, 21 determinants that impact an organization’s capability to implement BDPA were identified and validated. They were categorized into five determinant groups to form the basis for the BDPA Capability Assessment Tool (BDPA-CAT). Organizational Culture (DG5) was identified as the most important determinant group, followed by BDPA Operations (DG2), Data Characteristics (DG1), Technical Skills and Expertise (DG4) and Technology and Software (DG3). Analytical hierarchy process was used to assign weightages for the determinants. The BDPA-CAT was validated with four construction companies to reflect their respective BDPA capability levels, strengths and weaknesses.