If you've got a complex system -- a jet engine, for example -- the last thing you can afford is for it to fail at a crucial time. Frontier Technology says it has a way to predict those failures well before they can happen.
Frontier, whose top executives are based in Dayton, is commercializing a pattern recognition system called NormNet, which can analyze any system that uses sensors and predict when a component will fail, says Frontier senior scientist Sam Boykin.
Boykin says the technology began as a project with the Air Force Research Lab at Wright-Patterson Air Force Base. Along the way, the company has benefited from several Ohio Third Frontier grants and is nearing the end of a two-year, $350,000 commercialization grant.
Boykin says the technology has been successfully demonstrated to a number of companies, including Caterpillar, Parker Hannifin and General Electric Aviation. NormNet has also been tested at wind farm companies in Texas and on jet engines for the British Air Force.
"It's a software solution," Boykin says, and works "as long as there are some kind of sensors on the system. So when the system is operating healthy, the system creates relationships between each sensor and all the others. That's really how (NormNet) is able then to predict -- when it sees the degradation, it sees one of these relationships start to change."
He adds that "in all cases, we've been able to predict these failures -- sometimes days and weeks ahead of where they actually occurred."
Frontier's Dayton operations employ 20, including the addition of two to three new jobs over the past two years.
Source: Sam Boykin, Frontier Technology
Writer: Gene Monteith