How Systems Design Works

A systems designer works with managers to test structures and scenarios around those structures. The first step is to model the fundamental structure that managers wish to work with. Often it will look something like this.


This is a  simulation model of staffing requirements of a tissue typing lab for organ donations. It requires 24hr staffing by highly-qualified staff and rapid responses to donation situations.

SD Modelling has often been called “bath tub modelling”. The square boxes are called stocks (bathtubs) and pipes inbetween are flows. In this case, staff flow through the pipes and stay in the stocks while they are on leave or working in the labs etc. When the model is run, it is possible to see where all the staff are and how many are actually working, on leave etc.  it is then possible to get an idea of how many staff are required to keep the system functioning.

This staffing situation is also complicated by the fact that trainee staff (rookies) need to be supervised by experienced staff increasing demands on the staffing schedule. The pool of qualified staff are predominantly women so maternity becomes an issue in staffing mix. All these factors become important in the feedback dynamics of the model.

As a result, staffing mix and scheduling becomes a highly complex problem and a computer simulation is the only way of understanding the total number of staff required in each of the categories staff on leave, maternity staff et cetera.

Because it is absolutely critical that results from the tissue typing are provided in a timely fashion it is essential to maintain tissue typing information for the transplant teams.  As a consequence, The staffing of the lamps becomes a critical dynamic in the organ transplant process.

The design of this particular model was developed with the staff and management of the laboratory.  Using this process means that the management team will have confidence that their insights into the organisation and its process have been incorporated into the model and that they can trust the results and data that comes out of the model.