Waste Management Improves Route Efficiency

The Problem

Waste Management Inc. (WM), the nation's leading provider of comprehensive solid waste collection and disposal services for residential, commercial and industrial customers, had a monumental productivity challenge on its hands. Following a series of acquisitions of other trash-hauling companies, WM had a fleet of some 26,000 vehicles running nearly 20,000 daily routes. Many crisscrossed each other and were inefficient based on route length, distance between stops and other basic productivity indicators. WM was failing to reap the full economic benefits of its huge scale of operations. Yet the complexity of rationalizing its routes was enormous because of the large number of operational variables and frequent changes in customer requirements. While WM's most experienced and talented route planners could, with timely information, adjust routes to maximize their efficiency, WM's overall route management systems were not equal to the task, to the detriment of customer service and WM's profitability.

The Analytics Solution

WM sought out a new, user-friendly Web-based software solution and judged competing products based on their ability to meet three key goals: reducing the total number of routes, smoothing out total routes evenly over the course of a week, and adhering to business constraints (including vehicle capacity, route capacity, routing time limits per vehicle and time windows at the depot). The selected system, called “WasteRoute,” tightly integrated a geographical information system (GIS) with basic customer and route data in a relational database.

The full power of GIS in this analytics-based system is harnessed through the development of an origin-destination matrix, capturing the distance and time traversed between any two points factoring in speed limits, one-way streets, and other constraints. WasteRoute's underlying optimization engine also can organize routes according to individual customer criteria as well as geographic parameters. The integration of navigational data, an interactive map and customer information from WM's other systems, wrapped in a Web-based Java application, provides WM with a software solution that can be used by employees with relatively limited computer skills.

The Value

In its pilot phase in 2003, WasteRoute reduced the number of routes per location by an average of 10%, for a net saving of $8 million. Savings for 2004 were estimated at $44 million. Future savings are expected to result from route managers' increasing familiarization with the system and its deployment throughout the entire company.