Understanding improvisation in construction through agent-based modeling

Abstract

Improvisation is a spontaneous, impulsive, but rational decision-making process. Although unpopular, as it indicates deviation from plans, improvisation is unavoidable in construction as it helps address emerging non-standard issues or problems related to unforeseen uncertainties. Therefore, a construction project will benefit from applying an adaptive planning system that employs improvisation and reacts rapidly to unplanned incidents. This research study aims at advancing the knowledge of improvisation in construction through better understanding its dynamics. The study’s contribution to the industry lies in guiding construction planners and decision-makers to better manage the problems of unexpected uncertainty through enhancing the overall improvisational performance in a construction project. Agent-based modeling is used to develop a simulation model that depicts the improvisation practices at the level of planners working and interacting together in a construction project. The simulation model illustrates planner, project, and problem-related parameters that highly influence how each planner improvises. Moreover, the model considers different types of interactions that might shape the ongoing improvisational practices. Also, it identifies how different variations of improvisational parameters, related to the planners’ assortment and types of construction projects, influence the emergent improvisational outcome. Thereafter, Analytical Hierarchy Process survey is conducted among a group of construction experts to quantify certain behaviors and rules related to the model’s components. Also, the inputs of the developed model are validated while considering data from five large-size projects, located in a metropolitan capital. In addition, simulation experiments are conducted to develop linear regression models that anticipate the outcomes of the improvisational practices in a certain project from a given distribution or combination of improvisers. The established classification of improvise