A driver for McPherson Companies loads product at a company warehouse near Birmingham, Ala. Planning routes for the delivery of liquid bulk and packaged lubricants to customers across the Southeast is an extremely complex process, says Kevin Hodgkins, McPherson’s director of information technology and process improvement. By implementing a route optimization tool, company management plans to nearly automate its route planning, to reduce error and to increase asset utilization by as much as 25 percent, he says.
In trucking, nothing follows plan. At any moment, multiple load opportunities, driver emergencies, equipment breakdowns, weather, market pricing, changes in pickups and deliveries, and numerous other factors require dispatch to make quick but intelligent decisions. For very small carriers, these choices may be simple. But as carriers grow, so do the number of options for adjusting operations.
Humans can process a limited scope of information at one time, so most carriers use various decision support tools ranging from spreadsheets to dispatch software to highly sophisticated “optimization” systems that continuously gather data throughout an enterprise and optimize the balance as variables change.
McPherson Companies Inc., a Birmingham, Ala.-based distributor of industrial lubricants, hopes that such an advanced optimization tool will help it solve daunting routing challenges due to its product line, equipment types and deliveries, both scheduled and “on-demand.” For example, in order to deliver 100 orders in one day, managers at the 100-truck company might have to plan deliveries of 175 line items for vehicles and tank trailers with multiple bulk compartments.
“It is hard for the human head to get around the problem,” says Kevin Hodgkins, McPherson’s director of information technology and process improvement.
Optimization systems provide recommendations for improving operational efficiency, customer service and profits by finding the optimal balance among all variables. But a computer system is only as effective as the willingness and ability of humans to leverage the system’s capabilities. To get the full value from these systems, fleets often must make additional technology investments to ensure drivers and managers actually take advantage of their new tool.
Building profitable routes
For operations that require multiple stops per driver per day, route optimization is one of the most widely used types of optimization tools. By using advanced software to optimize daily routes, fleets such as Skokie, Ill.-based Peapod are able to schedule more orders per truck, reduce mileage and decrease overhead.
Throughout the 1990s, Peapod – an online grocery delivery company – used a manual process to plan daily delivery routes. Dispatchers would group orders by ZIP codes, says Thomas Parkinson, senior vice president and co-founder. As the company expanded its service into new markets, the routing process became decentralized and led to inconsistencies in planning. “We were figuring it out on the fly,” Parkinson says. “It was getting out of control.”
Five years ago, the company implemented a routing and scheduling tool from Descartes. Online orders from Peapod’s customers are injected into a route engine and turned into optimized delivery routes. Peapod’s warehouse management system, through integration with the route optimization tool, sequences pick orders for products, by order of delivery, for loading trucks at warehouses. The Descartes scheduling tool enabled an increase in orders per truck from 14 to 18, increasing the profitability of each order, Parkinson says.
Under the existing process, McPherson Companies’ enterprise system prints customer orders to its warehouses, depending on the location and type of product ordered. At some facilities, planning routes and dispatching trucks can be a full-time job for warehouse managers, and drivers typically sequence the deliveries by themselves, Hodgkins says. This month, the company plans to go live with A.Maze Routes, an optimization system from GEOCOMtms. Computer modeling of new delivery routes compared to current routes suggests that deliveries could be made with 25 percent fewer vehicles, he says.
Truckload optimization
In contrast to the route optimization tools designed for pickup and delivery operations, optimization systems for truckload carrier operations are based on a much different set of calculations. The goal for truckload operations is to maximize yield – a very complex calculation of the margin per load per day within the concept of the total network. Computation of yield depends on a number of variables, and carriers can have different definitions of what represents optimized yield.
Little Rock, Ark.-based Maverick Transportation uses several optimization products from Manhattan Associates, including Driver&Load, which provides Maverick’s customer service representatives (CSRs) with real-time recommendations for matching drivers with loads based on yield. The system gives the company flexibility to adjust the optimization calculations to provide recommendations that reflect its strategic goals. For example, a strategic goal might be improving the quality of life for the driver. Optimization that includes this as a factor will produce a different result from one based solely on maximizing revenue.
Maverick uses Driver&Load to optimize a driver’s time, says Doug Richey, vice president of sales and operations, Southern division. “It does a lot of calculations quickly, such as a driver’s waiting time, how long he has to wait for a load and how long it takes to make a delivery,” he says. The system also takes into consideration drivers’ hours-of-service requirements and helps the company get its drivers home on a regular basis. “We wanted an added tool to meet those needs of drivers.”
In addition to finding the optimal match of driver to loads, truckload optimization systems can help carriers quickly analyze freight and load offers to see how they fit within their network. Transport Industries – a 3,000-truck owner-operator carrier based in Mesquite, Texas – is using Integrated Decision Support Corp.’s Netwise optimization system to help create a new dedicated truckload operation it calls the Virtual Private Fleet. The fleet now consists of 10 trucks, but the company plans to reach 50 trucks this year, says Mike Voelk, director of information technology.
Through an exclusive agreement with a major third-party logistics firm, the company uses Netwise to “cherry-pick” from a wide variety of the 3PL’s loads via electronic data interchange.
“We get much more freight than we can actually handle,” Voelk says. Instead of analyzing the freight by its market price, the company analyzes the freight based on yield. The Netwise system dynamically analyzes the loads to create optimized “closed loop” trips for the fleet. Virtual Private Fleet offers cost savings to the customer based on improved asset utilization. The recommendations the software provides essentially are the best savings for the customer, Voelk says.
“As a dedicated carrier, (the market price) doesn’t make any difference to us. They will add more freight if they’re not saving money,” he says.
Human efficiency
Thanks to the flexibility of optimization systems to be customized to match a carrier’s unique operations and decision-making processes, fleets are able to increase office productivity.
“Our people were making very good decisions before we brought the optimization tool on,” says Maverick’s Richey. “This enables a lot of the thought process to happen for them – to enable them to make decisions quicker. As with a lot of software tools, we sell it to them as a ‘tool to help you do your job better.’ ”
The recommendations from Driver&Load appear in the load planning screen of Maverick’s dispatch system, Richey says. A CSR can see the optimal loads for each truck – up to five different choices ranked by score. “The key is that it optimizes it as quickly as possible,” Richey says. Before using the system, a CSR would have had to dig deep to find all the necessary information, such as a driver’s available hours.
“It has solved the everyday issue of having a lot to do, and time is money,” Richey says. “The quicker a driver gets dispatched, the quicker he gets to the customer, and so forth.”
McPherson Companies is hoping that its new system will help improve office productivity by decreasing the load planning process to no more than two hours, Hodgkins says. Managers should be able to spend more time on their other duties such as talking to drivers, handling emergency deliveries and going into the warehouse to help. And those efficiencies will be critical in keeping overhead – as well as equipment purchases – in check. “We are growing pretty rapidly,” Hodgkins says.
Matching recommendations and reality
On paper, the benefits produced by optimization systems may seem convincing, but in the end, the results only can be as reliable as the information provided to the system. Companies that implement optimization systems struggle the most with maintaining accurate data on drivers, tractors, etc., so that the system’s recommendations are something they can actually use.
“The solving of the problem is not the big prevention,” says Tom McLeod, president of McLeod Software. The key to improving the performance of optimization systems is data collection and validation of the data once it is added to the system, whether through EDI or manually, he says.
In addition to the data quality factor, “there is still the human element,” Richey says. Optimization systems may not take into account a detail that a dispatcher knows. By monitoring the percentage of recommendations accepted by each CSR, Maverick is able to improve its system continually by adjusting the tables in its software after finding out why certain recommendations were not accepted, Richey says.
Currently, Maverick’s CSRs accept the recommendations about 85 percent of the time. If the percentage drops, it most likely is a problem with the system, not resistance by a CSR, Richey says.
Another challenge for fleet operators is improving driver compliance to optimization systems. To track its driver productivity and route compliance, McPherson Companies plans to add a GPS handheld application to capture data in the field, such as driver hours, and a standard or baseline timetable for each route. “We see it as a driver training and discipline tool,” Hodgkins says.
Similarly, Peapod plans to increase route compliance by upgrading to Descartes’ next-generation routing and scheduling product. The company plans to use a system called Mobile Link, an application that runs on GPS-enabled cell phones. By knowing the real-time location of its vehicles, the company will be able to manage route compliance better and to dispatch drivers more dynamically, as opposed to creating one route per driver per day.
“That’s huge for us,” Parkinson says. “We’re looking forward to a successful upgrade.”
For companies that use advanced decision support systems, the real challenge is managerial, not technical. Getting employees and drivers to actually comply with “optimization” likely will require a lot of listening.
Ultimately, however, fleet executives must recognize that no major operation can – or should – be run by computer. Says Richey, “There is nothing out there that can make all decisions necessary to be made.”
Ignoring the market rate
Yield can be a strategic pricing tool
Because of the capacity shortages in the truckload market, many carriers have been successful in getting rate increases from their customers. After all, price is relative to supply and demand. But price is not an effective bargaining tool when compared to yield, says Stephanie Williamson, director of marketing for Dart Transit, a 2,200-truck carrier based in Eagan, Minn.
“Yield is sanitized,” Williamson says. “You don’t have to give away your secrets of what profit you are making. There is no way to back into a rate from sharing a yield number. It has too many variables in it.”
Last year, Dart Transit began using Netwise, a pricing tool from Integrated Decision Support Corp. that calculates the yield on every load – past, present or future, or any other way you slice it.
“(Yield) allows us to have a constant measurement,” Williamson says. “In transportation, when you are reviewing rates, you have the idiosyncrasies of headhaul and backhaul markets. In this pricing tool, the term of measurement is yield. A dollar in Dallas is the same in Chicago. Headhaul and backhaul – those things don’t exist.”
In negotiating rates with customers, Dart Transit can present a customized report from its Netwise system that says, for example, “here’s what yields well, and here’s what doesn’t yield well.” The company can show the customer why a particular load has a low yield: the length of haul, the shipper’s hours of operations, unloading time or deadhead. In short, yield explains why a particular load may not be a good fit for the carrier’s network, which helps Dart’s personnel explain to customers how to put in actions that are corrective, she says.
“Information is power,” Williamson says. “Either you can fix the rate, or it doesn’t matter how much money you pay – it is not going to work.”
