I couldn't find a better way to word my title but I will explain the situation.
I need to model collection routes for a single truck. Working time is from 08:00 - 12:00, then 13:00 - 15:00. Break is from 12:00 - 13:00. This is the case for three cities (three independent VRP analyses). There is always only one depot.
Initially I had problems getting VRP to function, so up until recently I have been generating routes manually with Route Analyst. I basically divided the collection points in rough groups that made sense geographically and that I estimated would fit in the 08 - 12 or 13 - 15 schedule, with the depot as starting and ending point. Of course, it took a lot of time and credits because I rarely got it right on the first try.
Last week, I finally figured out my problem with VRP and re-ran the routes I had already modelled. The results are very different, but not in the sense I was expecting.
For VRP, I created a "Routes" sheet with three identical pairs of the 08 - 12 and 13 - 15 working times, totalling 6 routes. They weren't always fully used but I wanted to make sure the programme didn't run out of options. Now, VRP is modelling routes in such a way that either only the 08 - 12 times are being used or the 13 - 15 times. In practise, it is giving me solutions in which 1. the truck is idle for part of its working day and 2. the collections last more days than necessary (which I concluded by comparing my Route Analysis results with VRP's).
I would like to prioritise routes in such a way that it first fits the collection points into the first pair of routes (08 - 12, 13 - 15, with break) and then the second and third pairs if necessary, but I can't find a field for something like that. Route names have to be unique and I can't assign two starting and ending times. In other words, VRP needs to understand that Routes 1 and 2 are a single working day and one part of the pair can't be ignored.
How would I go about solving this?
Thank you in advance!
