Waste Route Optimization Software Modeling Accuracy

Methodologies to Assess Waste Collection Route Times

In setting up the routing model’s productivity rates, there are two approaches:

1. Scale Data Calibration Method – Base the model on current route times derived from scale ticket data, which, in practice, has proven to be the most accurate method proving route times that are very close to actual times. This approach captures issues like traffic, maneuverability issues and other factors that are not available in the data.
2. Individual Customer Modeling – Base the model on a time to service each stop, plus a time to traverse the streets. Using the field-assessed time and motion studies, this method is still accurate.  However, it does miss more of the nuances in the route that affect routes time.  This approach uses the posted speed limits, a slow down and speed up time at a stop, and a pull time for each stop.

In aggregate, both approaches are accurate.  It is at the individual route level that the scale data calibration method has proven to be more accurate.

For the City of Virginia Beach in 2013, scale ticket data was summarized to assess the accuracy of the route optimization model using the Individual Customer Modeling. Because the routes were implemented in March of 2013 and drivers were not comfortable and proficient with the routes until June, scale ticket data from June and October was utilized for the analysis. Further, June through August were peak months, which follows the objective of the routing project to create optimized routes for peak months.  Exhibit 1 summarizes the data by collection day and by vehicle type. It demonstrates that the model estimates overstated the actual route times. For the tandem routes, the average route time from the Individual Customer Modeling was 0.825 hours longer than the actual time to complete the route, which exceeds the best practice of keeping actual within 15 minutes of the model.  For single axle routes, actual were 0.25 hours less than planned, which is an accurate assessment.