I have a paper submitted for this year’s Transportation Research Board annual meeting! There is still a long ways to go and perhaps it will not be selected, but there is some interesting things in the paper for users of TERM Lite. My paper is about modeling complex rehabilitation environments using TERM Lite. In my particular case, the issue is that my data set includes 25 different operators with a minimum of 25 possible different takes on how a bus should be maintained over its lifespan. Modeling those differences in TERM Lite doesn’t work easily be default, since TERM Lite was originally designed around the concept that it would be used at the local agency level. In this situation, there is most likely only one rehabilitation policy that applies to assets of a general type. For example, if most of your buses are of a particular make and model, then they probably get treated the same. There is one rehabilitation policy per asset type, which is exactly what TERM Lite expects and is designed to model. In this case, TERM Lite does a decent job.
However, when you want to assign two different policies to the same basic physical asset (for example, two stations in different parts of the city), then it gets tricky. Either you accept the single rehabilitation as “good enough” or create a work around. In my paper, I describe a method for assigning multiple rehabilitation policies to a base asset type that avoids many of the pitfalls that come with the most obvious workarounds, while creating a virtually infinite level of possible modeling detail. In my region, based on the format I’ve set and the number of agencies we have, TERM Lite goes from 600 possible rehabilitation policies for 600 asset types to over 1.4 million possible rehabilitation policies for the same 600.