Geography Reference
In-Depth Information
Figure 3. Layout of VRP Model and Inputs for Modeling Solid Waste Collection.
Given a polygon data-set, the user can specify the number of random
polygons to be selected.
The tool was applied to the parcel data to assign residential locations
(customers) on each of the four collection days to the 20 companies (Figure 4).
The randomness of the tool mimics reality in that there is no systematic
structure in residents choosing their solid waste collection company. All
residential locations in each of the four collection areas were assigned to a
particular company. No residential location is served by more than one
company.
The controlled collection scenario models six solid waste companies in
each of the four collection areas. Residential locations serving as solid waste
collection points are equally divided into six geographic clusters (Figure 5).
The purpose of this scenario is to give local companies with the larger number
of customers an equal share of customers. One of the concerns city officials
have about contracting solid waste collection to only one or two companies is
that approximately 18 companies would lose their customer base, thus hurting
local businesses. As a compromise, this scenario utilizes six companies with
residential locations clustered rather than spread across the city. Identifying
clusters in GIS can be performed using the Cluster and Outlier Analysis or Hot
Spot Analysis tools given a weighted variable for the data.
