In trucking, especially the fast-paced world of expedited trucking, the $64,000 question is: Where do you send your trucks? The answer is worth a lot more than $64,000. In fact, it's worth millions. The economics of trucking are about maximizing "loaded" miles and minimizing "empty" miles - which is easier said than done. "Our customers simply refuse to schedule their emergencies," cracks Joe Greulich.
Enter Dynamic Vehicle Allocation (DVA). A software outfit called Transport Dynamics Inc., which grew out of a research lab at Princeton University, has customized high-powered vehicle-optimization technology to help solve Roberts's scheduling problems. DVA uses three complicated algorithms - a demand-forecasting model, a fleet-management model, and a driver- scheduling model - to consider one simple question: Which truck should take this load?
To answer that question for each load, DVA considers every truck in the Roberts fleet - all 1,600 of them. It applies about 20 different rules and objectives to them, each factor weighted with a different point value: when the truck can pick up, how long the truck has been laying over, if its last haul was less than 75 miles, the number of miles the truck is from the pickup, and its "projected economic value" in the drop-off area.
The last figure is new for Roberts. It's based on where DVA predicts the company's business will most likely be over the next four days and what types of trucks will be needed. The software examines runs from the previous 365 days, then takes into account whether it's looking at a Sunday (the slowest day of the week) or a Friday (the busiest), at October (busy) or January (slow). Because each truck's status and location are constantly changing, so too is DVA changing, updating assignments based on current information.
"This is the biggest change we've ever made here," says Greg Mulhollen, manager of operations and planning. "It's bigger than the satellite. This is looking at how our business works."
A version of this article appeared in the September 1998 issue of Fast Company magazine.