What we have learned about Simulation

Our group has finished our Wiki subject perfectly. Here is our group work about Simulation, if you are interest in studying simulation, you may get some help from our works.

1. WHAT IS SIMULATION?

A simulation is basically the imitation of a real life event, process or circumstance. Simulations are used because they are usually an economically feasible way of experimenting what if scenarios of real events that couldn’t be done otherwise. The uses of simulations are very varied, from educational purposes like flight simulators to train pilots to weather predicting simulators which seek to forecast changes in the world’s climate by altering diverse parameters.

Models can be designed to explain simple situations like calculating sales commissions in a business, this type of models are called deterministic because they will always produce the same results when re-calculated. However, there are several instances where imitating real life events are considerably more complex due to uncertainty in the various parameters that influence the event.

2. MONTE CARLO SIMULATION

The technological advances undertaken over the years have greatly increased our ability to elaborate more powerful simulation models. The data processing power of computers makes complex simulations possible. One application of computers to simulation is the Monte Carlo simulation method which basically generates numerous random values for the parameters in a model to be evaluated, hence, turning the deterministic model into a stochastic model.

Moreover, this method is used for analyzing uncertainty propagation or how random variation affects the sensitivity of the model being evaluated. To effectively use this simulation method, a probability distribution that appropriately describes the data at hand must be selected. From this distribution, a computer draws random values which essentially is a sample from the distribution, thus this method is considered a sampling method.

Monte Carlo simulation can be easily implemented for models to determine demand forecasts, price variations, inflation effects on profits among other situations, using the following five steps:

1. Create a function that describes Y in terms of the variables identified as relevant.
2. Generate a sample of random values to use as inputs.
3. Evaluate the values generated in the model.
4. Repeat steps 2 and 3 as many times as desired.
5. Analyze the results using the appropriate tools like histograms for example. ( By Gerardo)

3. WHEN DO WE USE MONTE CARLO SILMULATION?

Whenever we need to make an estimate, forecast or decision where significant uncertainty exists, the Monte Carlo simulation could be the best choice for us. Otherwise, our estimates or forecasts could be deviation, as a result, the consequences would be inaccurate and the wrong consequence may adverse for our decisions. In the actual business environment, most business activities, plans and processes are too complex for an analytical solution. However, we can build a spreadsheet model that can give us numerical support to evaluate the plan. Therefore, if there are some changes in the parameters, we can get our results by changing numbers or by asking 'what if analysis.' This is straightforward if we have just one or two parameters to explore. Some business situations, however, involve uncertainty in many dimensions. For example, we are facing a variable market demand and we do not know about the competitors' plans, uncertainty in costs etc. If the situation sounds like this, we may find that the Monte Carlo method is surprisingly effective for us as well.（By Caesar)

4. OTHER USES OF SIMULATION IN BUSINESS DECISION MAKING

As stated above, simulation has been a very powerful tool to help managers make decisions in a business world of uncertainty and complexity. Another example of the use of simulation is in supply chain management (Kumar & Singla). The Distribution Network Simulator is a scenario-based decision making tool to help analyze the operational behavior of an existing or planned supply chain or distribution channel. The model uses entities such as nodes (factories, hubs, and other facilities), routes between the nodes, processes and policies, products and their demands, transportation modes (trucks, ships, and airplanes), shipments sizes, and service levels to build simple and complex model of supply chains and distribution networks. For example, it integrates the uncertainty in demand and transportation and uses a range of stocking and ordering policies to determine the fill rates. The Distribution Network Simulator is powerful in many aspects. For example, it helps to evaluate the effect of changes of costs in the supply chain, analyze the effectiveness of resource utilization or damages of bottleneck within the network, or the impacts of reduction in inventory or transportation costs.

Works Cited:

Baker, S. G. & Powell, K. R. (2009). Management Science: The art of modeling with spreadsheets. Hoboken, NJ: Wiley.

Kumar, R. & Singla, J. Simulation modeling in retail logistics and supply chain.

5. Simulation vs. Spreadsheet Analysis

Because I am the last contributor to the simulation wiki, I want to address a question that has not been covered: why use simulation over a spreadsheet analysis?

It is fairly simple to develop a spreadsheet analysis. However, as the analysis becomes more complex, there is a need for more complex computer based programs. Spreadsheets can perform advanced calculations to arrive at an answer, but it uses averages to get there. Because it uses averages, spreadsheets cannot accurately display the randomness and interdependence of the situation. However, simulation is able to anticipate the randomness and interdependence of the situation. Therefore, simulation is more applicable to the real world than spreadsheets. For instance, the time needed for assembling a widget may take an average of 10 minutes. However, a special order widget may take 45 minutes to complete. If you use a spreadsheet, you will have to use the average time, thus your analysis will not be able to accurately capture the variability that exists. Another difference between the two is that spreadsheets are static. This means that spreadsheets can only estimate a quantitative result for a single moment of time. Stimulation, on the other hand, is dynamic and can replicate your business reality. (By Kody)

Works Cited

"Justifying Simulation." Justifying Simulation: Why Use Simulation. ProModel, n.d. Web. 7 Nov 2010. .