PENGEMBANGAN MODEL PART TIME SERVER PADA SISTEM ANTRIAN

NUR INTAN ZUHAROH, Ir. Subagyo, Ph.D.

2016 | Skripsi | S1 TEKNIK INDUSTRI

The number of customer in a physical queue which usually exists in a bank or supermarket can become a factor for a consumer to decide whether he or she should join the queue system or not. The behavior of a costumer is determined by the perception of waiting time. This kind of perception is made based on the length of queue and the number of available server. In a perspective of queue length, the longer queue is, the higher willingness to leave the queue. This condition will make such a great lost for the company. To overcome this, the company should increase the quality of service and one thing that can be done is adding part time servers. Part time server is a server that will only be used to help the main servers to reduce a long queue and in resting time. After the queue has been normal, the part time server will be diminished. This experiment is conducted to develop a mathematical part time server model for decreasing the length of queue. The model is developed by using Monte Carlo simulation of the queue system in BNI 46 KC UGM. The factors that are considered in this model are server utility at earlier period, the length of queue at earlier period, the arrival rate at the present time, the interval of observation time, standard service time, and the average service time, and number of server. The model resulted from this experiment is able to decrease 63,64% the average number of customers in the system for Pattern 1, from 11 customers to 4 customers, with the end of period had 3 people in the server and 0 people in the queue, and 42,86% for Pattern 2 , from 7 customers to 4 customers with the end of period had 0 people in the system.

The number of customer in a physical queue which usually exists in a bank or supermarket can become a factor for a consumer to decide whether he or she should join the queue system or not. The behavior of a costumer is determined by the perception of waiting time. This kind of perception is made based on the length of queue and the number of available server. In a perspective of queue length, the longer queue is, the higher willingness to leave the queue. This condition will make such a great lost for the company. To overcome this, the company should increase the quality of service and one thing that can be done is adding part time servers. Part time server is a server that will only be used to help the main servers to reduce a long queue and in resting time. After the queue has been normal, the part time server will be diminished. This experiment is conducted to develop a mathematical part time server model for decreasing the length of queue. The model is developed by using Monte Carlo simulation of the queue system in BNI 46 KC UGM. The factors that are considered in this model are server utility at earlier period, the length of queue at earlier period, the arrival rate at the present time, the interval of observation time, standard service time, and the average service time, and number of server. The model resulted from this experiment is able to decrease 63,64% the average number of customers in the system for Pattern 1, from 11 customers to 4 customers, with the end of period had 3 people in the server and 0 people in the queue, and 42,86% for Pattern 2 , from 7 customers to 4 customers with the end of period had 0 people in the system.

Kata Kunci : the queue length, Monte Carlo simulation, part time server