The application of production planning approach on a large number of residential building construction project post natural disaster

After a large-scale natural disaster struck an urban area a large number of residential buildings construction is often needed. Such residential buildings are often built with more or less similar size and model therefore the construction is repetitive. The project management approach is normally employed in managing a large-scale residential building project. This study argues that due to its repetitive nature the construction of a large number of similar houses can also be viewed as a production system in addition to a project view. This study investigates the utilization of the production planning approach with the aid of a discrete event simulation method in managing the construction of a large number of residential buildings. This study simulates the construction of 500 similar houses. The study results show that the production management approach has the potential in managing a large-scale residential building project in addition to the project management approach.


Introduction
An earthquake with 7.4 magnitudes along with tsunami and liquefaction hit Palu City and its surrounding areas on 28 September 2018. The triple large-scale natural disaster caused the loss of thousands of lives and houses. After the disaster, thousands of houses needed to be reconstructed in many locations in Palu and its surrounding areas. The model and size of the houses in one location are often more or less similar, therefore the construction process is repetitive and can be compared with the conveyor belt system in manufacturing [1][2][3][4][5][6]. This study analyses the completion time of large-scale house construction when it is viewed as a manufacturing process.

Literature Review
There are two methods frequently used in planning namely activity-based planning and workflow planning [2,7].

Activity-based planning method
Construction planning in conventional construction project mainly focuses on the project's activities. This planning approach is called activity-based planning or project planning [7]. The critical path method is an example of an activity-based planning method [8].
In the activity-based planning method, the duration of each activity is a function of the work quantity and the production rate of the employed resources [9] as shown in Equation 1.
work quantity Activity duration production rate = (1) The activity-based planning method implies that the project completion time is solely determined by the summation of all the critical activity duration. Hence, the project completion time is only changed when its critical activity durations are changed [10].

Workflow-based planning method
The workflow-based planning method is the basis of production system planning which is commonly used in the manufacturing industry. The workflow-based planning method is also called production planning [11][12][13]. In the workflow-based planning method, the cycle time is a function of the work in process and the throughput as shown in Equation 2 which is also known as Little's Law [14].
Work in process The cycle time is the average time needed to finish one product, the work in process is the average number of products in the production system, and the throughput is the average output of the production system per time unit [14].
To analyze the correlation between the throughput, the work in process, and the cycle time in the construction of a large number of houses, this study uses the SIMUL8 application which is based on the discrete event simulation method. The discrete event simulation method is a simulation method that is commonly used in quantitative analysis of a construction process [15][16][17] 3 Definitions and assumptions

Definitions
In this study project duration is the completion time of 500 houses; the cycle time is the completion time of a single house; the throughput is the number of houses that can be completed per day; the work in process is the number of houses under construction; the interarrival time is the time between the commencement of each house.

Assumptions
In this study, it is assumed that all the houses have similar size and model; there is only one group of workers available for each activity; no parallel activities; and an activity duration is deterministic.

Model and simulation
Typical activities in single-story house construction withs their identifications and durations are shown in Table 1. Since this study assumes that there are no parallel activities, hence the construction starts with the foundation activity then followed by the wall, and so on as depicted in Figure 1.
By using the critical path method, it is determined that the completion time for a single house is 68 days. The single-story house construction sequence in Figure 1 is then simulated in SIMUL8 utilizing different inter-arrival times as depicted in Figure 2.

Figure 2
The simulation of single-story house construction in SIMUL8.

Results and discussion
The simulation result with various inter-arrival times ranging from twenty to ten days is shown in Table 2. The correlation between the work in process, the cycle time, and the throughput is shown in Figure 3. The simulation result in Table 2 shows that there is a negative correlation between the inter-arrival time and the work in process, that is the decrease in the inter-arrival time will increase the work in process.