What is System Dynamics?

System Dynamics is technique for describing and simulating the behaviour of complex socio-economic systems. It consists of four stages:

• Describing a systems material flow in terms of "stocks" (i.e. where the material is stored or built up) and "rates" (i.e. the speed at which the material moves in and out of the stocks)
• Detailing the information that controls the rates
• Capturing the feedbacks and time-delays that affect the system
• Calculating how the stocks, rates and information changes over time as the system is "run"

For systems where there are significant feedbacks & time-delays, where there are complex linkages between different types of materials (e.g. money affecting resources affecting behaviour) or where there are sudden shocks which can have significant knock-on effects, system dynamics can be more useful than other modelling techniques in:

• Understanding how the complex system behaves
• Predicting future performance or outcomes over time
• Providing support to decision making

Example: Simple manufacturing supply chain

Basic material flow: This simply shows the main "stocks" where the manufactured item is held and how it flows from one position to another. In this case material flows into the factory where it becomes "Work in Progress" after which it entrer the "Manufacturer Inventory". From there it is shipped to the "OEM inventory" (Original Equipment Manufacturer) by one of two routes: contract shipments or through a "spot-market" (e.g. commodity markets). From their it is used in a finished good of some kind (e.g. computer or washing machine).

Information flows: Each of the rates (the speed at which material is transferred between stocks) is controlled by some initial piece of information; production start rate (the amount of material entering the factory per day), the manufacturing cycle time (the time taken to complete an item out of the raw materials), maximum shipment rates etc.

Feedbacks & time-delays: These are the second-level pieces of information which drive the initial informatino flows. Often these are feedbacks from other parts of the system. For example, the production start rate is driven by a feedback from the manufacturer's inventory (i.e. if you have lots of inventory you might want to reduce how much you are making). In addition you have to take into account time-delays in making changes which mimic the normal delays in decisino making and implementation in any complex system (e.g. time to adjust inventory).

It is often this combination of material flows, with feedback and time-delays that leads to fluctuations in stock levels, supply, demand and ultimately performance.