Text

theses

The concept of supply chain management has been evolving in construction industry steadily since mid-1990s as more and more firms experience the power of collaboration for business growth and market expansion. Unlike most manufacturing supply chains, a construction supply chain is more unique and complex due to its characteristics such as one-off project, non-repetition, and highly-customized nature. Although many researchers have devoted their efforts in construction supply chain management and construction project management, there has been a significant gap between what are being studied and what are experienced in practices. The main contribution of this study to the academic literature is its effort toward filling this gap. This dissertation includes three essays, triggered by common operational challenges encountered in the construction practices. First, we study a resource-dependent project network consisting of multiple concurrent projects that are independent in operations but are subject to the same final quality inspection upon completion. In addition, each independent project consists of a set of activities, the earliest start times of which depend on the availability of required supplies and the precedence relationships among the activities. Thus, the supplier selection and coordination among these activities need to be considered to guarantee a timely completion of each project. The problem is modeled as a mixed integer linear program (MILP), which contributes the literature as a new mathematical model that describes a common operational problem emerged in coordinating the project resource allocation, the activity schedule, and the final project review operations. Secondly, we present a formal structural analysis of the problem introduced above, and investigate the effectiveness of a mathematical programming based heuristic. The heuristic decomposes the MILP model into three sub-problems aforementioned, which can be solved independently and consecutively. The overall objective is improved by updating a weight vector in the supplier selection sub-problem, where the shadow prices of some constraints in the project activity scheduling sub-problem and project review sequencing sub-problem are utilized to guide the weight updates. Our empirical studies show that the proposed heuristic is capable of generating near-optimal solutions in a short computational time. Finally, we conduct a study on project scheduling policies under uncertainty. We focus on the uncertainty that commonly exists in the duration of completing each project activity. Two problems are investigated in this essay. First, we study the impact of a collaboration policy on final project review, under both batch mode slot assignment and sequential mode slot assignment. The benefits of collaboration compared to its non-collaboration counterpart are shown using computer simulation. Then, the resource allocation problem based on time-resource tradeoff problem with uncertain activity durations is introduced. The general problem is discussed and an algorithm for a special case is presented. Five future extensions from this research are also discussed, such as (1) constrained project scheduling with renewable and nonrenewable resources; (2) project scheduling with inspection requirement and multiple objectives; (3) supplier selection based on 3PL and 4PL in practice; (4) sustainability application; and (5) alternative decomposition schemes of the heuristics.

ETD_7650

Ph.D.

Includes bibliographical references

by Zhengwei Wang

doi:10.7282/T3TH8Q1W

ETD doctoral

The author owns the copyright to this work.