Facilitated-mode model coding in simulation projects using the FaRe approach

A general problem for operational researchers is convincing stakeholders to implement the recommendations produced by our models. The literature highlights that this is a particularly notable problem in simulation projects in healthcare in the UK. One widely-acknowledged barrier is stakeholder engagement, with facilitated modes of modelling advanced as a remedy. Therefore researchers have been pursuing fully-facilitated modelling, i.e. involving the stakeholders in all phases of modelling. In discrete-event simulation (DES) the ‘model coding’ phase (building and open-box validation of the computer model) is a particular challenge, since converting conceptual models to working computer models is a time-consuming and technically-skilled task.
We have been experimenting with using the Business Process Model and Notation (BPMN) standard for flow mapping and DES modelling in hospitals in Italy and the UK. We called this the “FaRe” (facilitated and requisite) approach. Available through a number of free software packages, BPMN has many powerful features for this task. BPMN files can link with DES through standards such as BPSim or be imported into commercial DES software. Free software packages that incorporate both BPMN and BPSim and have commercial-level interfaces make it possible to pass through the model coding phase seamlessly. We will give an example of a successfully-facilitated Conceptual Modelling – Model Coding – Experimentation project in a fairly simple healthcare situation.
However, BPMN was not designed for this type of process mapping, and BPSim (and import to commercial DES packages) have some limitations. Consequently we have encountered technical barriers to using this tool for more-complex flows, as frequently encountered in modelling of inpatient flows. These arise principally from conceptual limitations with BPMN. We conclude therefore with suggestions and research directions for developing BPMN and its interfacing with DES to make this potentially very powerful tool fit to enable fully-facilitated modelling in a wider range of problem situations.