Since ancient times, billions of tons of topsoil, the nutrient-rich topmost soil layer essential for growing crops, has been lost from farms each year due to erosion.

One theoretical remedy, reasons Cornell graduate student and Iowa farmer Clay Mitchell, would be to recycle the eroded topsoil, which comes to rest in low-lying areas. But would this hypothetical undertaking be worth the trouble?

To get some answers, crop and soil sciences graduate student Mitchell, whose 2,000-acre corn and soybean farm is widely recognized for use of cutting-edge technologies, enlisted some unusual help -- Cornell master of engineering students in operations research and information engineering.

For their 2008-09 master's project, the students -- Shaan Qamar, Anshuman Bhairavbhat, Talha Omer and Kevin Ham -- set out to optimize the redistribution of topsoil over Mitchell's farmland. Devising efficient routes by which a tractor would pick up the topsoil, drive it back to the farm and dump it in the places most needed, the students calculated savings for Mitchell of roughly $6,000 to $10,000 a year.

"We were interested not only in where to take the soil and place it to gain maximum revenue, but also, in how to move it there," Qamar explained.

Dividing the farmland into a grid, the students started by noting the amount of harvest Mitchell collected from each grid square. Typically, a "sweet spot" of topsoil, about two inches in depth, leads to increased yield, Mitchell told them. Adding soil beyond that point produces diminishing returns.

By analyzing the relationship between topsoil depth and expected yield, the students deduced the actual topsoil depth at different parts of Mitchell's farm, and plotted tractor paths using optimization computer software.

But the tractor, when moving reclaimed soil to certain areas, would also compact the land as it drove, requiring the use of a plow afterward. The students had to reduce costs by using the tractor as little as possible, and therefore the plow. And to avoid wasting fuel and labor, they also had to consider which parts of the farm needed various amounts of topsoil.

The genesis of the collaboration was atypical for a master of engineering project, which placed second in the College of Engineering's Silent Hoist and Crane Company Materials Handling competition. Mitchell had described his idea for the topsoil redistribution in a paper for a systems engineering class taught by Huseyin Topaloglu, associate professor of operations research and project adviser. After reading the paper, Topaloglu decided to recruit some students to work on the problem.

Now graduated, the four master's students have handed over the reins to a new group of students this summer, who are tasked with figuring out exactly how the tractor would be programmed to follow the routes their predecessors have drawn.

Topaloglu, who noted that operations research is gaining recognition for use in such areas as financial engineering and health care planning, called the unusual nature of the project "refreshing."

"We are looking at something that's not really a traditional business problem, but it if pays off, it could make changes in people's lives," he said.