Comprehensive CP Optimization for Dynamic Scheduling in Construction

Abstract

Delays and cost overruns are common facts in construction projects due to its increasing complexity, the day-to-day dynamic changes, the stricter execution constraints, and the general lack of efficient scheduling tools to support the optimization of construction plans. Currently, many scheduling tools and techniques are available, in addition to a large body of literature that focus on schedule optimization. Such tools and techniques, however, do not adequately represent or incorporate various practical decisions and constraints, nor provide the project manager with the ability to examine the combinations of actions in order to either plan or bring the project back within the constraints. This research enhances the schedule optimization research by efficiently modeling real-life decisions and constraints, and develops a framework to optimize planning and corrective-action decisions; dynamically before and during construction. The development of the proposed framework starts with a basic model that suits the schedule optimization decisions at the preconstruction stage. This model is then extended to a generic model that accommodates the dynamic schedule optimization needs during construction. The enhancements and extensions are formulated in a generic mathematical formulation to optimize the schedule’s decisions at any stage. This formulation integrates a wide range of scheduling options (e.g., linear crashing, activity multimodes, overlapping, and multipath networks), and incorporates the project manager’s preferences about the corrective-action decisions’ implementation. The formulation also considers a variety of practical constraints (e.g., variable resource availability, correlated modes, and intermediate milestones); and uses a multi-objective optimization to tradeoff among the project time, cost, resources, and permissible schedule changes during construction. Based on the mathematical formulation, the proposed framework was then coded using the advanced v constraint programming tool “IBM ILOG CPLEX Optimization Studio”. To validate the model, multiple experiments on four case studies were used to prove the functionality, practicality, and its better representation of real-life construction challenges. Two of these case studies are taken from the literature to prove the ability of the comprehensive model to achieve better solutions. Construction experts were also consulted at multiple stages of this work to investigate the relevance of the framework. Introducing the proposed framework as an add-on to standard project management software is expected to change the practitioners’ perception that optimization is a theoretical and complex tool. Therefore, it helps to present optimization as a useful decision support tool for construction scheduling.