Organization: University of Pittsburgh

by M. Ulukus & G. Degirmenci

This case involves a real plant in Thailand in a high competitive market. With its commitment to please customers and improve delivering quality, customer satisfaction is one of its core values. At present day, the business has three manufacturing plants in Saraburi province. But, in this paper we only focus in one plant, which manufactures and distributes two types of cement called INSEE-Tong cement and INSEE-Dang cement. For easier understanding, in this paper these products are mentioned as premium and regular, respectively. Recently, as the domestic cement demand rose, the factory has encountered with long-waiting lines at the loading facilities. (the plants loading facilities are capable to serve pallet loading, bulk loading, and floor loading trucks; however, we only focused on pallet loading facility, as it is the easiest one to be modified.) This long-waiting time occurs severely from 4.00PM to 10.00PM, which is the peak time for incoming trucks to receive products. One approach to solve this problem is to reschedule incoming trucks. At present, the plant serves both customers’ pick-up trucks (PK) and the company’s delivery trucks (DL). As PK quite varies and is difficult to be scheduled, the rescheduling approach will not solve the problem effectively. Thus, our goal is to redesign the loading facility and change the process or policy so that overall waiting time in system is reduced. This will benefit the company in increasing its distribution capacity, increasing resource utilization in distribution facility, and gaining customers’ satisfaction.

Project Objectives

  • Using simulation software to help to study the dispatching process in the company's plant, to see the behavior of each work station and then find where the problem is.
  • To find out the factors or parameters are that affect the overall system.
  • To find the possible solutions for the problems that occur within the factory.

Summary and conclusions

By comparing results between the current and the proposed layout, we can see that the average time in system is decreased from 2.098 to 1.42 hours. Also, the number of waiting truck and the waiting time of server1, server2 and server3 is considerably decreased. Moreover, there is no information at server 4, that is the waiting number and waiting time is zero; in other word, it means that there is no queue built up at this server.

This might be a result of changing the assigned loading slots, as the new loading policy allows up to 3 slots for loading regular cement, and 3 slots for loading premium cement, comparing to the current loading policy, allowing 2 slots for each type of cement. However, we could not conclude that way until we prove that utilization at each loading station is increased due to increasing in number of services.

Moreover, the further experiment of weigh station policy demonstrated that the1 WI 1WO 1WIWO policy yield the lowest time in system. As a result, the proposed layout together with the interchangeable weigh station policy provide the optimal time in system of the trucks.