This study has been carried out for the course Planning and Design of Service Systems in a shopping center. The study relates Determining Employee Requirements which is a very important of planning step of designing service systems according to two points of view. First, it prevents long waiting queues in front of any service. Second it increases the efficiency of service providers.

Even if there are some analytical methods such as Waiting Lines Models to solve this type of problems, they are not adequate for representing the dynamic characteristics of the system. So the model is modeled by using simulation, one of the most effective management science tools.

As a result of this study, it is aimed to provide data to increase the utilization of the service providers, and also degrease the waiting times in queues by changing the number of service providers in each department. Because as a result of the simulation it is easy to see where to focus on. And also the simulation is flexible enough to answer what if questions. The study is modeled by using SERVICE MODEL - Student Version.

The problem takes place in a shopping center. It consists of five departments and a payment department. The departments are fruit & vegetable, meat counter, pastry, sport center and clothing or Dept1, Dept2, Dept3, Dept4 and Dept5 respectively as they will be called during this study.

The starting point is to evaluate the current system and consider the alternatives. In order to do that, some performance criteria are determined:

• Average utilization of service providers
• Average waiting time of customers in queues

• The system is a terminating one.
• Customers join a FIFO order queue if the service provider of that department is busy.
• All times are in seconds.
• All service providers of a certain department are identical.

In the system, there are five departments. In meat counter there are three service providers. And also in pastry department there are two service providers. In fruit & vegetable, sport center and clothing departments, only four, five and four customers can make shopping at the same time respectively. And there are three types of customers called cust1, cust2 and cust3 respectively. Also there are five service provider at the payment department.

As it was said there are there type of customers. The inter arrival times for customers is exponentially distributed with a mean of 40 seconds. They have been classified according to the shopping habits. These types are shown in Table1.

Table1. Customer Types

Additionally, customers have to visit the payment department which we can call Dept6.

To reflect the real world some situations are included in the model. These are:

• For a customer, it is given where to visit in Table1. But there is no obligation that he has to visit these departments in the given order. (He can visit the most wanted department first)
• A customer can visit a department that he has not thought to visit before entering the shopping center. (He can be attracted by a product or an ambiance of a department)
• A customer does not have to visit all these departments. (Sometimes he forgets to buy a product and leaves the shopping center)
• A customer can visit a department that he has visited before. (Sometimes he needs to buy two products from a department but he forgets. He buys one of them and goes to another department. But if he remembers he goes back to former department and buys the second product)

All these above are normal events that someone can do in a shopping center and defined with certain probabilities to reflect the real life to the model. To make this events possible, the probabilities for any customer to go from a department to another is shown in three consecutive tables.

Table2. Probabilities for Cust1 to go from a certain department to another one.

Table3. Probabilities for Cust2 to go from a certain department to another one.

Table4. Probabilities for Cust3 to go from a certain department to another one.

After visiting the departments, customers come to the payment department. There are five cashiers here and after paying, the customers can leave the shopping center. These cashiers have 20 minutes brake in every 3 hour. And the processing times for customers in each department are shown in Table5.

Table5. Processing times for customers in each department.

Also service providers working in meat counter, and pastry department have breaks. At meat counter, there two regular service providers. They have 30 minutes brake in every 4 hours which are successive. In this one hour period, a third service provider starts to give service. So a customer arriving meat department always finds two service providers here. But in pastry department, only one regular service provider exists. This provider has 20 minutes break in every 3 hours. 10 minutes before he leaves for break, a second service provider joins the pastry department. and this provider leaves 20 minutes after the first provider comes back.

For the other departments which have no service providers, certain number of customers can make shopping because of the space constraints. For example in front of fruit & vegetable department only four customer can stand. That means that only four people can make shopping. And this number becomes four and five for sport center and clothing departments respectively. But all these places are out of order for five minutes in every four hour because of arrangement and rearrangement .

The location of facilities in the shopping center is shown in Figure1. Also the flow chart of the system is shown in Figure2. Figure2 gives the logic of the working principle of the system. According to this logic, the model is developed.

In order to verify the model, following techniques were used:

• The model was observed with the animation option of Service Model to see whether the program runs as intended or not.
• The simulation was stopped at a certain point in time, and the results were examined.
• The program was introduced to other classmates to take their opinions.

An idealistic goal in validation is to ensure that the simulation model is good enough to be used to make decisions about the system. In fact a real shopping center is more complex than the one in the model of this study. But the main logic is the same. There are customers, departments, service providers, cashiers, downtimes for service providers and cashiers, random walk of customers, queues, complaints etc. The model in this study includes most of these subjects as sufficient as in real world.

The model is run after verifying and validating the model. As indicated before the model is a terminating one so the model is run for 12h. = 720min. = 43200sec. under predetermined conditions. And also sufficient replications of the model must be run. By this way the results becomes more reliable.

In this study, the current performance of a shopping center was evaluated in order to find the best solution according to employee requirements. First of all, input data were gathered together, and some assumptions were made to establish the simulation model. Then SERVICE MODEL Simulation Modeling Environment was used to model the system. Only one situation was evaluated to see where to focus and which improvements to be made.