Detecting changes in plasmas is compulsory for control and the detection of novelties. Moreover, automated novelty detection allows one to investigate large data sets to substantially enhance the efficiency of data mining approaches. To this end we introduce permutation entropy (PE) for the detection of changes in plasmas. PE is an information-theoretic complexity measure based in fluctuation analysis that quantifies the degree of randomness (resp. disorder, unpredictability) of the ordering of time series data. This method is computationally fast and robust against noise, which allows the evaluation of large data sets in an automated procedure. PE is applied on electron cyclotron emission and soft x-ray measurements in different Wendelstein 7-X low-iota configuration plasmas. A spontaneous transition to high core-electron temperature () was detected, as well as a localized low-coherent intermittent oscillation which ceased when increased in the transition. The results are validated with spectrogram analysis and provide evidence that a complexity measure such as PE is a method to support in-situ monitoring of plasma parameters and for novelty detection in plasma data. Moreover, the acceleration in processing time offers implementations of plasma-state-detection that provides results fast enough to induce control actions even during the experiment.