Decision Making Scenarios

To analyze the performance of the system during disturbances different scenarios will be studied. The purpose is to observe the behavior of supply chain in each scenario when appears a disruption which affect the normal manufacturing process in the production plant. In this situation the inventory maintained in the whole system will be utilized to meet the customers demand until back to the normal state of the system.

Variance in client’s behavior and distribution system also are going to be simulated. The plan horizon utilized in all scenarios to calculate average sales to market and inventory levels is a period of 5 weeks. To understand each scenario first is defined their three main characteristics: the System Design, Market and Client’s behavior.

13

The System design refers how compound is the supply chain, if it has a regional warehouse or central warehouse; will face a flat market behavior, and the third characteristic, clients’ behavior; where the scenarios will be under normal, good or subordinate clients. Each of the characteristics will be explained in detail.

System Design

In a Regional Warehouse system design, has can be observed in Figure 3 is assumed that the manufacturing facility is small, when products are finished they directly send it in batches of 6 units to the five regional warehouses; each one carry inventory of all products to supply the demand of 4 different retailers and each retailer sell the products to final consumers.

Due to manufacturing plant has no capacity to hold inventory, the decision of distribution from the plant to regional warehouses are made following these priority rules:

a. A region has priority if the inventory level of this particular region goes below the minimum inventory level.

b. Also the shipment for a region has priority if the distance from manufacturing plant is longer.

c. Depends of demand quantity of product (demand order) from each region.

To request the weekly amount of end goods, every five weeks is computed the new average weekly regional demand value; with this value and the current level of inventory in each region, the new weekly demand order is made.

The average weekly regional demand value is calculated at the beginning of each period (as mentioned before, one period is 5 weeks), by using the last period demand value and the average sales to clients of the last 5 weeks in a weight ratio of 2:1.

Using the average weekly demand value, the weekly decision of regional demand will follow these rules:

14

a. If the current inventory level is lower than 0.66 of the maximum level, the weekly demand will be 1.2 of the average weekly demand value.

b. If the current inventory level more than 1.7 of the maximum level, the weekly demand will be zero.

c. If the current inventory level is more than 1.2 and less than 1.7 of the maximum level, the weekly demand will be 0.75 of the average weekly demand value.

Since regional demands not have set a delivery time, during the period the weekly demands are accumulated.

In the second type of system design, the Central Warehouse, the manufacturing plant has a central warehouse where all the finished products will be storage until it is received weekly demand from regions. Adding the central warehouse in the plant will create a buffer to handle the market fluctuations. With this kind of design the minimum and maximum inventory levels in the regional warehouses will drop. The Figure 4 shows the Central Warehouse System Design.

Figure 4. Central Warehouse System Design Source: Made by the author

The distribution of products from central warehouse to regional warehouses will depends on the current inventory level, which is calculated at the beginning of each period based on the average sales to clients. That means the regional warehouses will not send demand orders every week to the manufacturing plant,

15

instead, the plant will transfer the necessary amount of products to maintain the maximum inventory level.

Market Behavior

The market behavior presented in all scenarios is flat market. In a flat market the levels of sales to clients remains almost the same during all periods (6 units per product per retailer).

The average weekly sales of the system is 6 units per product per retailer, giving a total weekly average sales of 720 units (6 products * 6 units/product * 20 retailer), the region weekly average sales is 144 units (6 products * 6 units/product * 4 retailer) and product weekly average sales 120 units/product (6 units * 20 retailer).

Clients Behavior

In the different situations the system is set also by the clients’ behavior. Each one is compound of a set of rules which retailers based their decisions at the moment to send orders to regional warehouses. Those clients will be normal, good or subordinate. Under normal client’s behavior, retailers support their order decisions on forecast and experience, while their inventory levels are fixed: maximum of 60 units per product (which covers 10 weeks of average sales) and a minimum inventory level of 24 units per product (4 weeks of average sales). In this case, for the regional warehouse the minimum inventory level is 96 units, which represents the minimum level of one retailer 24 units multiplied by 4 retailers in each region, and a maximum inventory of 240 units/product (60 units *4retailers).

Retailers send orders to regional warehouses each week and this orders need to be fulfilled within one week. To calculate the weekly orders, first for a period of 5 weeks the retailers calculate the average sales value. This average sale value is calculated by old average value and the sales of the previous 5 weeks with a weight ratio of 2:1. After calculate this value, retailers make their weekly orders decisions considering the following situations:

16

a) The new order quantity is 0, if the current inventory level is more than 1.7 of the maximum inventory level.

b) If the current inventory level is more than 1.2 of the max level, the new order value will be 75% of the average sales values.

c) If the sum of the current inventory level, open orders and one fixed order is less than minimum inventory level, a special order is issued to return to the minimum level.

d) If the overall inventory is less than 66% of the maximum the order will cover the difference in 2 periods.

For good clients the average sales value for each period is calculated from the old average sales value and last plan horizon with a ratio of 5:1. The inventory levels at retailers and regional warehouse represents 6 weeks of average sales. The maximum and minimum inventory levels will change if there a case of missed to market.

For weekly orders the quantity requested is the difference from maximum level of inventory from current inventory and inventory on transit. Also during the week urgent orders are issued because of the following situations:

-The sum of current inventory level and the inventory in transit below the minimum inventory level an urgent order is issued to get back to the minimum level.

-There is a miss to market.

In subordinate clients’ situation, the average sales to market are calculated in the same way as normal clients, the weight ratio between old average sales value and the last period sales under this behavior is 5:1.

The inventory levels on retailers are calculated based on the average sales, for maximum inventory level is 6 times the averages sales. But the max-min inventory level is not fixed, will vary depends on weekly inventory levels. If more than twice in a period of 5 weeks, the inventory level drops under minimum level, the maximum and minimum levels will increase 10% per time. Instead if inventory level is more than 3 times average sales, the maximum and minimum levels will decrease 5%. The scenario with subordinate clients is based on daily decisions. The orders

17

Warehouse Flat Normal Good Subordinate

1 X X X future demand the end products are pushed to regional warehouse and then they are in charge to supply the demand from retailers which is also based on forecast.

Scenario 2 and 3 represents a pull-push based systems, Scenario 2 still is push system because maintain the regional warehouse system design, but also is pull because under good clients the inventory level are more realistic and just carry 6 weeks of inventory. For scenario #3, the system will be added the central warehouse and the products are storage there before they are sent to regions (pull system); but also is push because of based their order quantity decisions on normal clients’

behavior (forecast).

Scenario 4 is the pull based system because their design includes the central warehouse design which maintain the products in the plant until is requested in each region and also because the decisions are made daily based on previous day consumption.

18