Abstract for: Agile SD: a fast, effective, reliable approach to model development

This workshop demonstrated the Agile SD process and gave participants experience of its use both a facilitated white-board approach to capturing a quantified picture of an issue, and the same software model. The method specifies the time-path of the issue of concern. Performance aims (reducing crime rates, for example) are always traceable to stock-quantities (criminals) unless they are themselves stocks (number of criminals). The process therefore tracks the causes of the problem-issue until one or more stocks are reached, validating that causality with estimated time-charts for each item. The behaviour of the stocks is then explained by adding flow-rates (new people taking to crime, or giving up), again supported with time-values for each flow. Finally, the causes of each flow are traced back to three main root-causes decisions by actors in the system (e.g. penalties for crime), external factors (e.g. unemployment) or existing stocks (e.g. police numbers). This step identifies interdependence and feedback among the stocks. The workshop both showed the approach in use for a real case and applied the process to an example from the workshop participants, to show how it can be done from a standing start, including live development of a working model.