A supply chain is a coordinated system of organizations, people, activities, information and resources involved in moving a product or service in physical or virtual manner from supplier to customer. Supply chain activities transform raw materials and components into a finished product that is delivered to the end customer (Anna Nagurney: Supply Chain Network Economics: Dynamics of Prices, Flows, and Profits, Edward Elgar Publishing, 2006, ISBN 1-84542-916-80).
I most cases managing a supply chain is a complex operation. Even if the demand stays predictable and steady, the timing and speed of the movement of material and or information throughout a complex supply chain is a very complicated matter and may require very elaborate IT infrastructure and very clearly understood policies. The picture, however, becomes even more challenging in the face of an unpredictable demand or long delays at some parts of the chain.
When the demand changes in an unpredictable way, it sends a disturbance up and down the supply chain creating oscillations that grow as you move downstream from the point of disturbance. These oscillations result in large and costly inventory or long backlogs leading to the loss of customers.
A products/components or services flow through an interconnected chain of entities/firms, they accumulate in a variety of stocks till they are moved to the next stock. The policies or process rules determine the flows into and out of any stock. An important aspect of a supply chain is the fact that it usually involves more than one organization. Each organization (or a division inside a large organization) devises its own decision rules and policies to maximize profitability and return on investment. The decisions made by any organization, however, feeds back into other organizations and in turn influence their decisions.
System dynamics is ideally suited for modeling complex supply chains. Using system dynamics modeling, one can experiment with various policies and see their impact on reducing the oscillations and the bullwhip effect in the face of variations in demand and/pr unpredictable delays at various locations.
A simple supply chain may be composed of a retail store, a distributor, a factory wholesale and the factory. The factory turns the raw material into finished goods and delivers them to a wholesale business which in turns may provide the products to distribution centers that will them provide them to the retail stores. The customer order comes through the retail store. This issue was analyzed back in 1961 by the founder of the system dynamics field, Jay Forrester. He demonstrated the instability of such a supply chain in terms of demand amplification, inventory swings, bullwhip effect, etc.
At each location in the supply chain there is some sort of inventory. The outflow from the inventory is shipment to the next station in the supply chain or to the end user. The input for the stock of inventory is all the previous orders that have been placed from the upstream station and are in transit. This is the supply line. A simple, one station depiction of the system dynamic model is shown below:
Orders come into the stock of backlog. As shipment is made, the stock of backlog is depleted. The orders is placed from the backlog into the supply line. The items then move from the supply line into the inventory. Defining an variable called “Effective Inventory” the results of the simulation for a 60 week period is seen below.
The observed behavior is derived from the decision rule for placing the orders. In this model, the goal is to compensate for any shortfall in the desired inventory and compensate for the backlog. In addition, we add the amount shipped and reduce the order by the amount in the backlog. The current policy as formulated in the model creates a backlog of items in the steady state.
This is a very simplified model. In practice, there are many inventory and backlog stocks along the supply chain and there are many feedback loops throughout the system.
MapSys is well suited for modeling very complex supply chains in order to design effective policies that will reduce the cost and improves efficiency.
|
