The three-year, $3 million Department of Defense grant is aimed at reducing excessive vibration--a maintenance issue for all helicopters--and increasing the readiness levels of U.S. helicopter squadrons. If successful, the project could also increase the cost efficiency of helicopter maintenance.

"We've equipped Apache and Blackhawk helicopters at McEntire Air National Guard base (near Columbia) with sensors that pick up an array of frequencies associated with vibration," said Victor Giurgiutiu, a mechanical engineering professor who is joined on the project by department colleagues Abdel Bayoumi and Bill Ranson.

"By keeping detailed logs of the helicopters' maintenance and correlating those with data from the sensors, we hope to establish a 'cause and effect' understanding of when certain parts require adjustment or replacement," he said. "That might mean that some parts need to be replaced more frequently and others less so."

Such maintenance issues might seem routine, but helicopters often are grounded when needed parts aren't available immediately. With the ability to better predict when parts are wearing out or need adjustment, the military could keep more helicopters flying and prevent many catastrophic failures. When fully deployed, such systems could save hundreds of millions of dollars in operation and support costs.

The 18 sensors mounted on each helicopter are gathering reams of data that must be interpreted and compared with written maintenance records and onboard observations from the helicopter pilots. To handle the task, the research team is using a powerful data-mining machine that was donated to the College of Engineering and Information Technology by NCR and Walmart. With 400 gigabytes of memory and 1.6 terrabytes of hard drive storage, the data-mining machine has the power to track and sort myriad data from each helicopter.

"We're the first to get physical data from the helicopters and use that information to predict premature failure of parts," Bayoumi said. "The Army can't use anecdotal data in making maintenance decisions on helicopters; they have to have statistical proof."

Information from the project is being shared with other military researchers and similar research could be used on other military vehicles, including tanks and other motorized equipment.

12/01