It breaks away from the traditional fix-after-failure maintenance process by finding small, hard to identify problem before major failures occur.