Reliable, fail-safe electronics are crucial to the future vision of self-driving cars. Researchers at the Oberpfaffenhofen Training and Analysis Center are involved in a basic research project to develop a hitherto unique fail-operational approach for autonomous vehicles. Online diagnosis is to be used to detect incipient errors (degradation processes) in an electronic/electrical system early on (before the error actually occurs). In order to realize this novel anticipatory error detection method, the research team is working on a physical model of the cause and impact of errors in electronic systems. The specific focus is on energy wiring systems, i.e. supply lines and power distributors.