Inventory is both an asset and a liability. It exists at all stages of a supply chain. The primary purpose of these inventories is to buffer the uncertainty arising from demand, process and supply. Therefore, simulation technique should be used for supply chain inventory planning. This paper discusses the inventory planning of a supply chain through simulation, which consists of a manufacturer, distributor and retailer. The supply chain inventory model is developed and simulated in the Arena simulation tool with an aim to minimize the total system-wide costs which consist of production cost, inventory holding cost, transportation cost and stock-out cost. It helps in determining the production rate at the level of the manufacturer, and the Re-order Point and order quantity at the distributor and retailer levels.
Simulation
is one of the most popular of all quantitative techniques,
because it can be applied to operational problems that are
difficult to model and solve analytically. It is a technique
to mimic the operations of various kinds of real world facilities
or processes, usually on a computer with appropriate software
(Law and Kelton, 2000 and Kelton, Sadowski R P and Sadowski
D A, 2002). Determining the objective of a simulation is a
starting point for the process. This is followed by a set
of steps to develop the simulation model and eventually the
model is run to generate some results (Schniederjans, 1999).
The steps of the simulation process are as follows: definition
of the problem, formulation of the model, validation of the
model, generation of the results after simulation runs, analysis
of the results and decision-making based on the analysis.
Simulations can be expressed in software using conventional
programming languages. However, the most cost-effective approach
is usually to use a commercial simulation system like Arena
5.0. These systems include graphical tools for building models,
automated routines for testing them under different conditions,
and reporting tools for analyzing the results.
Simulators
incorporate variability quite naturally which is an important
advantage because sales, shipments, prices and countless other
aspects vary quite a bit in real world supply chains and this
has a major impact on how the supply chains perform. The simulation
models handle the variability by using proper distribution.
Although simulations are better than mathematical models at
exploring the effects of variability, they are not as good
at finding optimal solutions. The best way to handle this
in simulation is to vary the value of one or more parameters
in a systematic way and look for the one that gives the best
fit. |
