FAYETTEVILLE, Ark. - Long-distance collaboration can be a challenge for faculty researchers, and it is rare for graduate students. For the past year, Julie Watson, a graduate student in industrial engineering at the University of Arkansas, has worked with Helen Benton of Aston University in England on a project sponsored by the National Science Foundation to determine appropriate supply chain management tools for the new Internet economy.

With the changes in logistics and manufacturing, most businesses rely on other companies to supply raw materials, parts, transportation and a host of other goods and services necessary to deliver the product to the customer. When this supply chain is more agile in responding to requests, a company can meet the customer needs more effectively.

Watson and Benton are looking at the best processes for determining the planning and control system for supply chain flexibility. Management systems can be either centralized or decentralized. Each has advantages and drawbacks, but making the wrong choice can cost a company millions of dollars.

"A responsive supply chain is crucial for businesses if they are going to meet consumer demands for goods," explained Watson. "But the planning and control system will determine if information, materials and money flow efficiently."

Businesses have tried a variety of approaches. The development of the Internet has allowed increased speed of information transfer, but it has also heightened customer expectations. Although businesses have many options, each one requires a sizeable investment of time and money, which makes it important for companies to make the best choice.

Enterprise Resource Planning (ERP) systems were initially considered the solution, according to Watson. ERP implementation has also been marked by dramatic and costly failures, such as Mobile Europe and Dow Chemical.

"ERP survey reports show that only 33 percent of respondents found improved customer responsiveness," said Watson. "And only 57 percent thought that ERP actually improved the visibility of information."

Watson and Benton set out to address this situation by identifying a set of measurements that companies can use to determine what type of control system is most appropriate for their company. By applying a selective set of metrics across a supply chain, simulation techniques can be used to determine the best system for that business.

"It is important that the metrics used be applied throughout the supply chain," Watson explained. "That is the only way to make a fair and accurate comparison of an entire supply chain under particular industrial conditions."

The researchers also found that it was important that the metrics be cross-functional and based on processes rather than functions. They identified six key processes: order fulfillment, customer service, transportation/distribution, warehousing/storage, planning/forecasting/scheduling and information capability.

"It is important that the metrics be balanced," said Watson. "Logistics managers tend to focus on logistics measurements, for example. But it is important not to get fixated with metrics from one area at the expense of others. Paying too much attention to costs and ignoring customer satisfaction could cause their market share to decrease."

Watson and Benton built on research from The Logistics Institute at the University of Arkansas, which identified 98 possible measurements. From these they developed the Integrated Metrics Framework, which uses less than two dozen metrics to determine the best supply chain control system for a given company.

For a given company, centralized and decentralized systems can be compared for various conditions by using rich simulation. This produces a large amount of data that can be plugged into the essential metrics to determine which type of system best suits that company under those conditions.

They will present their findings at the Fourth Annual Managing Innovative Manufacturing Conference in Aston, England.