Defects Prediction Towards Efficincy Gains in Construction Projects

Abstract

One of the construction industry’s major problem areas is the potential negative impact of overlapping tasks within on-site work programmes as part of an integrated critical-path, work-breakdown-structured approach; this work seeks identification and mitigation of error and defect resulting from concurrent work-packages during the construction phase. A main challenge addressed is an accurate means and method to predict, at the early pre-construction planning stage, the cumulative defects that can result from overlapping tasks during construction, and subsequently seek to avoid, mitigate and better manage overlapping tasks that influence negatively project cost and duration. This paper presents work at the initial preparatory stages of a research project that seeks identification and understanding of the behaviour of overlapping task variables that result in error and/or rework; future work shall seek ultimately to go towards development of a predictive model to address cumulative defect. The research work being conducted is argued as significant in that it seeks an increase in the reliability of overlapping task completions during the construction phase. Objectives include: identification (and distribution) of defects during the construction phase, towards future; defect susceptibility for specific activities; identification of key overlapping task(s) variables; and, examination of the behaviour of each variable and its impact upon the overall project quality in general. Initial findings (where a percentage method determines rate of each factor for project phase and the source that caused the defects) indicate that the construction phase has the highest rate of defect appearance (67%) followed by the design phase (30%). Project process causes most defects (43%) followed by people causing 38% of defects.