In previous work we have introduced and discussed the execution monitoring paradigm Ontological Control (OC). The fundamental idea of OC is that the execution of a controller for a hybrid system should be monitored and that discrepancies between expected behavior (according to some model) and real behavior should be detected and classified as being either due to external actions (disturbances) or due to violations of ontological assumptions (faulty expectations or inadequate model).In this paper we report a first set of experimental results where the OC theory has been implemented and tested on data from a real process control system: The ABB STRESSOMETER rolling mill flatness control system.