Education: Materials handling university

Published: Sep 8, 2008 1:00:00 AM
Media Contact: ,

Source: Modern Materials Handling

The future of materials handling, logistics and the supply chain may be shaped by the research that is happening now at colleges and universities.

By Bob Trebilcock, Editor at Large

Let's face it: Business moves at warp speed today. While the best systems integrators, equipment manufacturers and software companies value research and development, the fact is, most have to focus on the problem at hand right now. Spending five or six years to investigate the stresses on cartons on a pallet stacked three high on the floor is a luxury they can't afford.

But taking the time to investigate a problem until the bitter end, even if that investigation leads to a dead end, is exactly what academic materials handling and supply chain programs at leading colleges and universities bring to the table.

"For every good idea, a researcher has a number of not so good and even bad ideas," says Kevin Gue, an associate professor at Auburn University. "But since we don't answer to Wall Street every quarter, we can ask the question: Why does it have to be like that? Ask it often enough and you might get an answer that leads to innovation."

Here's a look at some of tomorrow's questions being asked today at leading supply chain academic programs around the country.

Walter Zinn
professor of logistics
Ohio State University

One of the issues capturing Zinn's attention today is catastrophic supply chain risk. Safety stock, Zinn says, will overcome the everyday risk of a supplier not delivering critical parts to keep a line running. But, what happens when there's a long-term strike at a port, or a distribution center or a factory burns to the ground, floods, or is destroyed by a hurricane? To answer that question, Zinn turned to the same companies that simulate what-if scenarios to determine risk for insurance companies.

"Insurance companies hire these companies to tell them the risk of a terrorist attack on any building in the U.S.," says Zinn. "Why not use that same technology to assess the risk to a warehouse?"

While an insurance company evaluates risk according to its financial exposure, Zinn's research will provide supply chain managers with a model to determine the risk of losing a facility and then ways to mitigate that risk.

"This will allow supply chain managers to make decisions as complex as redesigning the supply chain to make a facility less important, or take steps, like reinforcing the building, to reduce the risk of losing it." Why does this matter? "It's become important because we've spent the last 20 or 30 years making supply chains lean," says Zinn. "We used to have safety stock lying around. As we've reduced costs, and we've been successful doing that, there is no slack. The consequences are more severe."

Kevin Gue
associate professor
Auburn University

According to Gue, his favorite question as a researcher is: Why does it have to be done that way? He began asking that question after visiting a warehouse in California. "The facility is a 24/7 operation that attempts to deliver orders to customers as fast as possible," says Gue. "Their constraint is that the last FedEx truck leaves at 5:00 p.m. Watching the operation, we found that 5% of their orders for the day were completed in the 30 minutes after the last truck left." Gue and his students began investigating whether a warehouse can reallocate workers at particular times of the day to improve the service performance, as opposed to the throughput, of the order fulfillment system.

"It turns out that with some fairly sophisticated analytical tools that we've just developed, a warehouse can develop policies that can improve its service performance by reallocating workers," says Gue. "We're building decision-support tools to complement a warehouse management or labor management system that would be focused just on that problem—getting more customers their packages sooner. Our early findings suggest that if done properly, we can game the system to do that."

Larry Lapide
director of demand management
MIT Center for Transportation and Logistics

Several years ago, Lapide changed hats and left AMR Research for MIT's Center for Transportation and Logistics to launch Supply Chain 2020, a project described as "a multi-year research effort to identify and analyze the factors that are critical to the success of future supply chains." "Academics often look behind them and analyze what happened," Lapide explains. "At MIT, we wanted to get ahead of the curve and look at what successful supply chain management is going to look like 15 years out."

Lapide and his colleagues spent the first 18 months creating a definition for supply chain excellence, which is a supply chain that is strategically aligned with the corporate competitive strategy. Next, they identified the macro factors and trends that they believe will contribute most to supply chains of the future, including issues like the aging of developed countries, the maturing of emerging economies, employing the unemployable and managing supply chains in light of rising energy prices. Finally, they addressed global issues, including the rise of Asia and other global trading blocks. "What we're doing now is modeling three different worlds in the future and asking what might happen to a supply chain under those scenarios and what can we do about that today," says Lapide.

Marshall S. White
professor emeritus
Center for Unit Load Design at Virginia Tech

For years, White and his students have researched what goes into designing the most efficient pallet and unit load. Today, the Center for Unit Load Design is taking that concept one step further into something White calls systems-based design. They look at the interactions of packaging, pallets and unit-load handling equipment as product moves through a facility or the supply chain. The idea is that too often materials handling systems, industrial packaging and pallets are designed in a vacuum, without looking at how the specifications of one component might impact the performance of another.

Currently, one of White's students is researching how compression stresses are distributed across a pallet and the product on that pallet, especially when stacked several high on the floor or in a truck. "What we find is that it's not consistent and varies tremendously from load to load," says White. "The more we can understand and change that distribution by looking at the design of the packaging and the pallet, the more we can reduce the cost of a unit load." How long does it take to answer a question like that? One of White's students has spent the past four years mapping compressive strength distribution using pressure sensitive film that records the stress distribution across the deck of a pallet as a function of pallet and package design.

Using those results, the student is creating mathematical models that will allow the industry to manipulate pallet design and reduce the cost of distribution packages. "By the time the project is done, she'll have a Ph.D.," says White. "It'll have taken about eight years of research before it's valuable to the marketplace. This is the kind of research industry should be doing, but doesn't have the time to do."

Bob Graves
Krehbiel professor of emerging technologies
Thayer School of Engineering, Dartmouth College

Automation control systems are at the heart of every materials handling system. The control software that sits between an execution system and automated materials handling equipment determines how efficiently pick modules are replenished, orders are picked, and totes are routed to packing and shipping areas. Compared to the cost of an automated storage and retrieval system or new conveyor, changes to the software are cheap.

"If you can get 30 to 35% improvement with software changes, you're addressing the most cost-effective way to improve throughput," says Graves. For the last five years, Graves and one of his Ph.D. students have been putting a new concept in controls software, known as Holonic Controls Architecture, to the test at Dartmouth.

The idea is to see how the concept, which Graves first heard about in the 1990s, can lead to improvements in throughput, uptime and accuracy over traditional control architecture. Working with a real-world system description, control code, system simulator and test order data provided by Vanderlande Industries, B.V., Graves has been able to complete head-to-head testing with a rebuilt system against the original control code.

The result: "We were able to demonstrate throughput improvements of up to 35% with a revised control system versus a traditional control system," says Graves. More importantly, Graves thinks the project is an example of how the materials handling industry and the academic world can successfully work together to innovate.