We propose a new methodology for creating embedded software that meets real-time deadlines. Our approach is a synthesis of real-time analysis and traditional systems debugging techniques, based on two main ideas. First, we use probabilistic modeling techniques to make quantitative predictions about system behavior, as opposed to real-time analysis techniques whose results are binary and often p...
