The report describes the activities carried out to identify new indicators for assessing Power Quality disturbances in medium-voltage distribution networks. Higher-Order Statistics (HOS) have been applied to detect and validate voltage sags recorded by the QuEEN monitoring system. Among the higher-order moments, variance acts as the effective value in detecting events, while skewness and kurtosis help determine the validity of the event. A distribution network model was also validated through simulation by comparing the voltage signals of events provided by QuEEN with the simulation results. The exploratory activity comparing the performance of traditional voltage sag indices and Cluster indices, based on monitoring data, showed that the N₃b index and the Cluster B index, potential “candidates” for regulatory action, are actually equivalent if probable constraints on the minimum duration of “to-be-regulated” events are adopted. It was also found that using classifiers based on Deep Learning techniques for identifying false voltage sags, alongside the active criterion in QuEEN, has a significant impact on the accuracy of index evaluation, which generally improves. This assumption was confirmed by the correlation activity between voltage sags and protection intervention data provided by UNARETI. It was verified that events classified as false or undefined by QuEEN, but with a duration greater than 90-100 ms, are mostly to be considered true. The report includes a summary of the findings from the literature review on the use of probability distributions for assessing disturbance levels in networks. Finally, it describes the initial updates made to the QuEEN site (visualization of Clusters, classifications based on AT/MT origin, etc.) and the results of the feasibility analysis conducted for implementing Machine Learning and Deep Learning applications in QuEEN developed within the research.