Abstract: The present invention relates to methods and systems for safety and/or performance monitoring of an Automated Driving System (ADS). The method comprises obtaining a plurality of Performance Indicators (PIs) generated by each of a plurality of ADS-equipped vehicles based one or more driving sessions, wherein the plurality of PIs are of at least one PI-type. The method further comprises modelling each PI-type of the obtained PIs based on an Extreme Value Theory (EVT) model, wherein each modelled PI-type is indicative of a probability for exceeding a PI-value for that PI-type, and evaluating each modelled PI-type against a corresponding predefined requirement. Then, if the evaluation fails, determining that the ADS has violated one or more safety or quality requirements, and transmitting a first signal indicative of the violated one or more safety or quality requirements to an ADS management system and/or a second signal to the plurality of ADS-equipped vehicles.