Enhancing Construction Project Performance Through Integrated and Optimized Supply Chain Management ^†

Abstract

The construction industry is shifting towards integration, digitization, and automation, necessitating an adaptive logistics system for enhanced performance. Despite this shift, poor supplier performance and lack of collaboration among stakeholders often cause delays and cost overruns. This research addresses these issues by proposing an optimization model for the planning phase of construction projects. The generalized mixed-integer linear programming (MILP) model is developed that optimizes supplier and process selection of construction projects, demonstrated through a numerical study. Results indicate that optimizing these decisions in the planning phase can significantly improve supply chain performance, enabling better cost and time management for construction projects traditional or modular.

1. Introduction

Construction Supply Chain Management (CSCM) often comes with challenges such as a lack of collaboration among construction project stakeholders and a lack of process integration which results in inefficiencies and cost overrun of a project [1,2,3]. The information generated by various sources of CSC during the planning phase contributes to fragmentation, which eventually leads to significant negative performance impacts, resulting in increased project completion time and cost [4]. A recent study [5] did a comprehensive review of modular CSC and highlighted research gaps. They identified that on the production level, there is a need for a more collaborative and integrated planning system that involves manufacturing modules’ companies and logistics companies to improve information sharing and enable collaboration among project stakeholders. And on the supply chain level, decisions related to resource allocation, production, transportation, supplier selection, and inventory planning need to be integrated to reach a mutually beneficial SC for its stakeholders. This calls for advanced mathematical optimization solutions to improve integration and collaboration in modular CSC planning. Another recent study highlighted [6] that there is a need to develop a CSC model that integrates logistics, production, and construction onsite phases for understanding and diagnosing supply chain performance. One recent study [2] emphasized the need for integrated thinking to achieve value-oriented goals in CSCM. Based on the above research gaps identified in the literature, the following question can be framed that needs a solution: How can a mathematical model improve the performance and collaboration of a CSC by integrating SCM decisions in the early planning phase of a project?

Therefore, this research aims to improve the performance of a construction project by developing a mathematical mixed-integer linear programming model that will consist of variables and parameters of SC processes required for planning a project that will help the management of a project to collaboratively utilize this model to select an optimized combination of modular activities and their suppliers in terms of cost and time.

2. Research Method

The hierarchy of project management usually differs depending on the nature of the project, the country, and the budget. In this paper, a simple residential building project is considered with the following stakeholders as represented in Figure 1.

Considered Stakeholders

The client is the project owner who finances the project. The General Contractor (GC) is the focal person managing the project from start to finish, ensuring timely delivery and client satisfaction. The GC also collaborates with the designer, evaluates suppliers using a mathematical model, hires them for required activities, and oversees their management. Additionally, the GC hires a sub-contractor to manage lower-level operations. The designer is responsible for the building’s modular configuration and the processes required for its construction. Before the GC evaluates suppliers, the designer provides the modular configuration and technical details to the GC. Element suppliers provide materials to process suppliers for module manufacturing and coordinate with both process suppliers and the GC. Process suppliers, who handle project activities from sub-assembly to final building assembly, coordinate with material suppliers for material availability and with the sub-contractor for timely availability of modules, equipment, and labor. They also report overall progress to the GC. The sub-contractor, hired by the GC, manages lower-level project operations, monitors, and schedules activity completion, and follows up with process suppliers on their work.

3. Mathematical Model

The model focuses on the optimization of the construction project at the planning level and then at the operational level. The supply chain starts when the contractor wins the construction project bid from the client. Then he receives design specifications from the designer and will use this developed model to evaluate material and modules suppliers and allocate resources to execute the project. The number of activities required and the required configuration of the building are already finalized before selecting suppliers. The following figure illustrates the participants and their information flow.

3.1. Mathematical Model Details

The developed mathematical model is the mixed-integer linear programming (MILP) model. The following are the details of the model. The constraints are not shown due to space limit:

3.1.1. Model Assumptions

• A process can have only one active successor process.
• A process can only produce one element.
• The mobile parts must be produced by someone other than the main contractor of the project.
• Material supplier only provides one material resource.
• The raw materials are available at all times in the right quantities.
• They do not produce a surplus.
• The elements are transported in shipments whose sizes are fixed before.

3.1.2. Sets

• $P$ Production process set,
• $S$ Building elements set,
• $Smob\in S$ Transportable elements set,
• $Sfix\in S$ Non-transferable elements set,
• $I$ Supplier set,
• $T(Days)$ Project period of time set.
• $\mathrm{Pr}ojectLimit(Days)$ 100

3.1.3. Parameters

• $T_{sp}$ Transition matrix,
• $T_{TotalTime}$ Project duration,
• $T_{tr}$ Transportation time needed from supplier I to the supplier J,
• $T_{tr Max}$ Maximum transportation time,
• $C_{pi}$ Supplier’s production cost,
• $C_{tr}$ Supplier’s transportation cost,
• $Q_{prod(p)}$ The quantity of elements to be produced,
• $Nee{d}_{(s,p)}$ The quantity of elements needed to produce one process,
• $\mathrm{Pr}oductivit{y}_{(i,p)}$ The productivity of a supplier for a specified process,
• $\mathrm{Pr}oductio{n}_{(s,p)}$ The quantity of elements produced by a process,
• $Lot{s}_s$ The shipment size of an element,
• $Ca{p}_{(i,p,t)}$ The capacity of a supplier to do a specific process at any time.

3.1.4. Objective Function

The objective function of minimization is composed of two elements, which are the process cost $C_1$ and the transportation cost $C_2$.

$$C_1={\displaystyle \sum _{p\in P}}\hspace{1em}{\displaystyle \sum _{i\in I}}\hspace{1em}{\displaystyle \sum _{t\in T}\left[C_{pi}\times X_{pit}\right]}$$

(1)

$$C_2={\displaystyle \sum _{s\in Smob}}\hspace{1em}{\displaystyle \sum _{i\in I}}\hspace{1em}{\displaystyle \sum _{t\in T}\left[E_{(s,i,j,t)}\times C_{tr}\right]}\forall j\in I,i\ne j$$

(2)

4. Numerical Example

For the application of this model, a case study is developed consisting of five suppliers that can do all the functions ranging from material supply for modular making to final assembly building. To test the robustness of the model, a total of 18 activities are required to make this final modular floor building. Activity 1 is the final assembly activity combining two floors, while activity 18 is the material supply activity to module makers. Activities from 2–17 are module making processes. Each floor requires eight activities shown in two different colors. The model can also select and optimize the required number of activities to do a project. In this numerical example, all the activities are required, so all will be selected by the model. The construction supply chain (CSC) network for the case study is shown in Figure 2.

5. Results

We tested the following conditions as shown in

Table 1. Result summary of the supply chain configuration scenarios.

Supplier	Objective Function ($)	Shipments Sent (Truck)	Total Processes Completion Time (Days)
With only 1	153,835	81	86
With All	122,828	55	39
Without 2	139,870	71	78
Without 4	122,868	56	46
Without 1	Infeasible solution
Without 1 and 2	Infeasible solution

with this model, which are with only supplier “1” (pre-determined supplier based on experience), all suppliers (free choice of suppliers), without supplier ”2”, without supplier “4”, without supplier “1”, and without supplier “1 and 2”. The results of these scenarios are also shown in the same table.

This table provides a comprehensive analysis of the impact of different supplier configurations on the objective function cost, number of shipments, and total process completion time. When utilizing only Supplier 1, the objective function cost is $153,835, with 81 shipments sent and a total process completion time of 86 days. This scenario demonstrates the inefficiency and higher costs associated with relying on a single supplier. In contrast, utilizing all suppliers results in a significantly reduced objective function cost of $122,828, 55 shipments, and a total process completion time of 39 days. This configuration highlights the efficiency gains and cost savings achieved through a diversified supplier base. Excluding Supplier 2 increases the objective function cost to $139,870, with 71 shipments and a total process completion time of 78 days, indicating the significant role Supplier 2 plays in maintaining lower costs and higher efficiency. Excluding Supplier 4, the objective function cost is $122,868, with 56 shipments and a total process completion time of 46 days. This scenario shows a slight increase in cost and completion time compared to utilizing all suppliers, suggesting that while Supplier 4 is beneficial, its absence does not critically disrupt operations. Notably, the absence of Supplier 1 alone, or in combination with Supplier 2, leads to infeasible solutions, indicating the indispensable role of Supplier 1 in the supply chain. These infeasibilities underscore the critical dependency on Supplier 1 for the viability of the supply chain operations.

Overall, these results highlight the substantial benefits of employing a diversified supplier strategy to optimize costs, reduce the number of shipments, and minimize the total process completion time. The results also emphasize the necessity of certain key suppliers, particularly Supplier 1, whose absence can lead to infeasible operational scenarios. Thus, by using this model, project managers can analyze valuable insights for their project supply chain management and evaluate the importance of strategic supplier selection and diversification. Therefore, it can be concluded that collaborating with multiple suppliers is better than sticking with only 1 supplier, and this model can improve decision-making at the planning level and can reveal with which supplier the most collaboration and coordination is required depending upon the importance. In this way, the model can help in achieving better collaboration and coordination with important suppliers. Since this model is for planning-level users such as general contractors, designers, and sub-contractors, they can collaborate in choosing suppliers and processes by running this model in different scenarios and conditions like it is performed in this paper to achieve better decision-making beneficial to stakeholders.

In addition to this, the model can give useful insights about the completion time of each process with respect to each scenario. In Figure 3, it can be clearly seen that the condition with all supplier choices clearly performs better than other conditions. The CSC network of the best scenario with all suppliers is shown in Figure 4.

6. Conclusions

The presented research introduces a mixed-integer linear programming (MILP) model designed for selecting construction suppliers and determining the number of activities required to complete a building, as specified in a transition matrix as input. However, in this paper, only the supplier selection example is shown due to space limits. This model incorporates various real-world parameters pertinent to supplier selection problems and employs a novel approach with the transition matrix to optimize the process selection based on given data. The research’s practical application is illustrated using a two-floor building where all processes were necessary, demonstrating the model’s ability to optimize based on the transition matrix and other parameters. The significance of this research lies in its potential use by general contractors and project managers during the early planning stages to optimize project activities and supplier selection. It also facilitates collaboration among logistics functions, enabling the evaluation of different supplier and activity scenarios to improve project operations and decision-making. Additionally, the model’s consideration of shipment times enhances understanding of suppliers’ operational performance. This addresses gaps identified in previous studies regarding the lack of supply chain planning in pre-construction phases and underscores the benefits of advanced mathematical optimization for integrated supply chain planning in modular construction.

Despite its strengths, the model has limitations, such as only optimizing production and transportation costs, and not accounting for factors like environmental impacts or logistical assumptions. Future research should incorporate these aspects and consider the uncertainty constraints to enhance the model’s applicability in the execution phase.