Modelling Workflow

Overview

A deterioation modelling project is a complex process involving many files and parameters that interact in various ways. Often, you may struggle to ‘tame your model’. That is, your model may give counter-intuitive outcomes or - more commonly - fails to respond to changes in constraints or inputs in the way you think it should.

As you can see from this page, the Cassandra Default Road Network model involves more than 200 different options, thresholds and constants which can be tuned to influence the model in various ways. This can be intimidating and confusing to modellers, leaving them stranded with no fixed idea on how to calibrate and adjust their model to accurately reflect the behaviour of the network they are dealing with.

In this section, we try to address this challenge by proposing workflows for modellers to use in their deterioration modelling projects. We give specific emphasis on how and when to alter model parameters and which parameters influence which aspects of model outcomes.

The workflows discussed below assume that all model setup files as well as the model input file is in place and that the model runs without error. For more information about these setup files, please refer to the ‘Guide’ section in the Juno Cassandra Framework documentation.

Validation Workflow

Once your model is up and running, the first step is to assess the reasonableness of your model’s predicted deterioration rates. See this page for a suggested workflow for model validating and adjustment with tips on what parameters to change to increase or decrease rates of deterioration.

Modelling Workflow

Once you are confident that your model provides a reasonably accurate representation of your network deterioration, you can proceed to evaluating budget scenarios with treatment allocation and optimisation done by Juno Cassandra. See this page for a suggested workflow.

Troubleshooting

Often you may run into a frustrating situation where, no matter what you do, your model does not want to behave in a sensitible way. In such a case, you should step back and drill down into details to see if there are some underlying problems with either Increments, Resets or Treatment Triggering. See this page for some pointers on how to troubleshoot your model.