OTAWA uses static analysis performed at machine level combined with hardware models to provide sound overestimation of WCET.