Abstract

Uncertainty is an inherent characteristic of construction projects. Neglecting uncertainties associated with different input parameters in the planning stage could well lead to misleading and/or unachievable project schedules. Many attempts have been made in the past to account for uncertainty during planning for construction projects and many tools and techniques were presented to facilitate modelling of such uncertainty. Some of the presented techniques are widely accepted and used frequently like Project Evaluation and Review Technique (PERT) and Monte Carlo Simulation, while others are more complicated and less popular, such as fuzzy set-based scheduling. Although accounting for uncertainty has been a topic of interest for more than four decades, it was rarely attempted to account for uncertainty when scheduling repetitive construction projects. Repetitive projects impose an additional challenge to the already complicated construction scheduling process that accounts for the need to maintain crew work continuity throughout project execution. This special characteristic necessitates producing scheduling techniques specifically suited to resource driven scheduling.
Therefore, the main objective of this research is to produce a comprehensive scheduling, monitoring and control methodology for repetitive construction projects that is capable of accounting for uncertainties in various input parameters, while allowing for optimized acceleration and time-cost trade-off analysis. The proposed methodology encompasses three integrated models; Optimized Scheduling and Buffering Model, Monitoring and Dynamic Rescheduling Model and Acceleration Model. The first model presents an optimization technique that accounts for uncertainty in input parameters. It employs a modified dynamic programming technique that utilizes fuzzy set theory to model uncertainties. This model includes a schedule defuzzification tool and a buffering tool. The defuzzification tool converts the optimized fuzzy schedule into a deterministic one, and the buffering tool utilizes user’s required level of confidence in the produced schedule to build and insert time buffers, thus providing protection against anticipated delays affecting the project. The Monitoring and Dynamic Rescheduling Model capitalizes on the repetitive nature of these projects, by using actual progress on site to reduce uncertainty in the remaining part of the schedule. This model also tracks project progress through comparing the actual buffer consumption to the planned buffer consumption. The Acceleration Model presents an iterative unit based optimized acceleration procedure. It comprises a modified algorithm for identifying critical units of the project to accelerate. This model presents queuing criteria that accounts for uncertainty in additional cost of acceleration and for contractor’s judgment in relation to prioritizing critical units for acceleration. Moreover, this model offers six strategies for schedule acceleration and maintains crew work continuity.
Together, the three developed models offer an integrated system that is capable of accounting for uncertainty in different variables through different project stages, aiming at helping managers keep repetitive construction projects on track. The presented optimization technique is automated in an Object Oriented program; coded in C# programming language. A number of case studies are analyzed and presented to demonstrate and validate the capabilities and features of the presented methodology.