Optimizing Transportation Planning Through Simulation

Optimizing Transportation Planning Through Simulation

The value of analytics and optimization tools for managing supply chains depends on the underlying data. The data, it turns out, could be a lot better. 

Statistical averages are the data on which most transportation planning and execution is based. Logistics organizations use statistical averages as a basis for transportation policy because they’re easy to comprehend and calculate, and provide a fixed numeric value that is easily input into optimization tools.

The problem is that statistical averages are fundamentally unreliable. They completely ignore the inherent variability that underlies nearly every aspect of transportation, said Mike Mulqueen, senior director of product management for Manhattan Associates, a global supply chain technology provider. 

“From order sizes, to lane volumes, to travel times to fuel prices, using a single, discrete value to represent each of these variables during strategic planning will lead to a poorly designed transportation policy that will ultimately be reflected as inefficiencies in daily planning,” Mulqueen said.

Statistical averages can distort reality. If half of a shipper’s orders are 5,000 pounds and the other half are 40,000 pounds, the average shipment is 22,500 pounds, an amount that isn’t even close to the actual weight of the shipments. 

See all ARO Logistics-related content.

Enter simulation modeling, which considers external variables such as lanes, modes, rates, carriers and regulations, but also internal variables such as new products, suppliers or warehouse locations.

For example, when converting freight from prepaid to collect, a long-term strategic decision with cost implications, key elements that must be modeled for variability include order quantities and product mix; order quantities of neighboring suppliers; freight rates; fuel prices; order frequency; and transport lead times.

Daily planning doesn’t require simulated modeling, because variables such as order quantities, rates and capacity are known at the time of execution. It’s the strategic planning process that should be informed by calculated probabilities, rather than what Mulqueen describes as the “forecast certainty” implied by fixed statistical averages. “You have to account for known variables or you risk optimizing on data that doesn’t reflect what could happen in the future,” he said.

Manhattan Associates is planning a 2013 rollout of a transportation simulation and modeling tool. The company currently provides simulation modeling as a service to a handful of customers. One current customer, Papa John’s Pizza, is using simulation to optimize transportation planning for servicing franchises.

The 2013 product release likely will be integrated with Manhattan’s transportation lifecycle management suite, or on top of Manhattan’s supply chain platforms. 

Supply chain leaders such as Home Depot use simulation to constantly challenge conventional wisdom and the status quo. Home Depot uses Manhattan’s transportation management system, and Mulqueen hopes the retail giant will become one of Manhattan’s first customers for its upcoming simulation tool.

“In a daily planning tool, the cards are for the most part dealt to you,” he said. “With this type of tool, the potential savings are much greater because you can impact overall transportation policy.”

Manhattan in October formed a new unit, Manhattan Mobility Labs, to focus on mobile applications and extending them into the Manhattan Associates platform of supply chain optimization solutions. New services from Manhattan Mobility Labs will leverage the Manhattan’s cloud-based platforms.

Contact David Biederman at inexdb@comcast.net.