To Nick Fahey, area manager of the family-owned Southern Oregon Sanitation hauling firm in Grants Pass, Ore., the clunky term “route optimization” breaks down to some very simple concepts, the cornerstone of which is making more money without skimping on customer service.

“To me, it's about how we can do a route faster, pick up for more customers in a driver's day, reduce the overall mileage for the fleet, and reduce the trips to the landfill/transfer facility,” Fahey says. “I want to be as profitable for the company as possible without impacting the customer financially. Because, the more efficient we are as a company, the longer we can put off a rate increase for our customers — and that keeps them happy.”

For Southern Oregon that's important, as it's a third-generation family business that's been operating in Oregon's Josephine and Jackson counties for over 60 years, serving both commercial and residential customers.

“'Route optimization' involves the creation of the least amount of routes to service customer stops and sequencing those stops in the most efficient order to minimize travel time or distance,” says Ewe-Leng Lim, chief knowledge officer at the Institute of Information Technology (IIT), a route optimization software provider based in Magnolia, Texas. “It means servicing more customers with the least number of vehicles and minimizing driving time and distances while meeting all business constraints.”

It can even involve such details as when drivers should leave for the landfill, when they should resume their collection routes and lunch break considerations, Lim says.

Making It Work

Southern Oregon's Fahey points out that a refuse fleet needs to do its homework before it can install a route optimization system and use it properly. Route optimization also involves spending cash — in Southern Oregon's case, roughly $20,000, he says.

Fahey's firm used two software packages — FleetRoute, a product of Germany's CIVIX LLC, and ArcView, a geographic information system (GIS) developed by Redlands, Calif.-based ESRI — to begin the process of building more efficient routes. But, the firm needed to gather some information before it could start using the packages.

The information included:

  • City and county maps

  • Customers' global positioning satellite (GPS) map points

  • What size of container each customer has

  • The average weight of trash in the container

  • How often a customer is serviced

  • Where the fleet's trucks are housed

  • What transfer site or landfill the vehicles use

  • The maximum tonnage of refuse that a particular truck can hold

  • How long it takes to unload a particular truck at a transfer facility or landfill

  • Street segment speeds and lengths

  • Length of the drivers' lunch breaks

  • How much time it takes for pre- and post-trip truck inspections

  • How long the fleet expects that driver's day to be

“All of these variables need to be input into the computer program to get appropriate routes,” Fahey explains. “Once we had this data, we then began manipulating it using the [software].”

Karl Terrey, product manager for ESRI's ArcLogistics software suite, notes that it typically requires several days worth of consulting and implementation before a route optimization system can go “live.”

“The flow of data must be carefully examined so that the correct information is available for the route optimization application,” Terrey says. “Once the data to be routed is made available, it must be ‘geo-coded’ to a specific side of a street segment and placed at the location of the pickup. Many systems will include a ‘geo-coding’ mechanism as a part of the application.”

Once the data and mechanisms to get it into the optimization software are validated, Terrey says the average person needs about one to three days to be trained on the system's workflow and functions. “Usually, it is all ‘back office’ personnel doing this work — allowing drivers to keep carrying out their days' work as usual.”

In Southern Oregon's case, Fahey says it took two weeks to set up the map appropriately. “For us there was a steep learning curve,” he says. “It took us at least a month to figure everything out to make the program work correctly with the customer data.”