Parametric construction scheduling

Joseph John Orczyk, Purdue University

Abstract

Building owners, construction managers, and contractors need a scheduling model that quickly and economically develops a construction schedule from the conceptual information. This dissertation examines the use of multiple linear regression and computer simulation as methods to produce construction schedules. The methodology of this research consisted of polling 54 construction experts to rank project milestones and parameters by importance. Values of the highest-rated milestones and parameters from actual projects were collected. This data was then used to demonstrate how regression and simulation could be used to develop the models. There are four conclusions warranted from this research. The first and second conclusions are that owners, designers, and large and small builders from all regions of the United States generally agree upon lists of both standard project milestones and standard project parameters. Third, the demonstrated results show that regression analysis is useful for describing the relationship between the parameters and the timing of the standard project milestones. And fourth, computer simulation is not as useful to describe this relationship.

Degree

Ph.D.

Advisors

Chang, Purdue University.

Subject Area

Civil engineering

COinS