Pilots have for years been able to test emergency procedures and decision making skills through the aid of flight simulators, whereas managers have been told to learn from experience. In the complex, dynamic world managers operate in this is no longer acceptable. It is simply impossible to learn quickly enough, or to effectively apply that learning to new situations due to the complexity.

System Dynamic modelling for the first time provides managers with a similar tool to the pilot's flight simulator.

It is now possible to develop a computer simulation model of complex problems and then test new decisions, structural changes, or strategy through the simulator to obtain insights into their future impact on the organisation.

Two important principles of system dynamics are: -

a. All decision/policy actions occur within feedback loops.

b. System structure determines system behaviour

A simple inventory system where the Order Rate is determined from the gap between the Desired Inventory and the current Inventory produces an inventory graph that increases steeply at first, then builds slowly to the desired inventory level as shown above.

However, if we change this system structure by including Goods on Order and a Receiving Delay then the behaviour of the system changes completely as shown in the following diagram.

The delays introduced into the system structure by the Goods on Order and the Receiving Delay cause the Inventory level to overshoot the Desired Inventory, then continues to oscillate as the Inventory strives to meet the Desired Inventory level.

In complex, dynamic systems this behaviour cannot be estimated or accurately determined by any other method than by System Dynamic Modelling.

The true power of this process allows management to simulate the real state of their organisation due to the ability to include the "human element" that is completely missed when normal estimating or forecasting is used. Economic forecasting without including morale, esteem, culture, employee attitude in the equation does not provide a clear indication of future outcomes.

System Dynamic Modelling allows for these important elements to be included, and therefore produces system behaviour that is more closely matched to "real world" outcomes.