Skip to main content Skip to navigation

Introduction to Simulation for Variance Reduction

Simulation is an activity that allows one to draw conclusions about the behaviour of a real (or proposed) process by studying the characteristics of a model.  It is used for evaluating the trade-offs between performance and resource requirements and to determine the “optimal” design.  It is used when the process or system is complex or when the risk of failure of the real system is high.  Simulation models can also be physical, such as cockpit simulators for training airline pilots. These are physical “mock-ups” of a real airplane. There is also a computer component of the model which represents conditions and creates the “emergencies”. This type of model is also used for training plant operators.

Simulation models provide quantitative performance measures of a given system configuration. They allow us to see the gain in performance for each addition or reallocation of a resource.

Since a simulation model is a computer model, and not an actual trial with real life customers, bottlenecks and unforeseen failures that cause long delays, poor service, bad product, or even dangerous conditions do not result in disappointed, disgruntled, or injured customers or employees.  For example: If your one of your design alternatives is to build a 10,000 sq. ft. warehouse, you can “test” it out without ever hammering a single nail. No capital investment is required.

There is no shut down of production nor any use of raw material with a simulation model.

The turn around time from concept to results is a matter of minutes, whereas in real-life it could be months or years. 

In most cases, simulation is the best way of testing a high level design at the system level. However, it may be necessary to test one or more elements in greater detail, possibly because these elements contribute to the risks and vulnerability of the design, or because the characteristics of performance are not easily quantifiable. In such cases, prototyping is a good approach to test a design before detailed decisions are made.

Further material will be presented as part of the course manual.