His recent research focuses on time series and data stream mining. On time series, he proposed the first time series distance invariant to complexity and speed-up methods to compare massive amounts of time series data under warping. On data streams, he has worked with classification with label latency and proposed efficient unsupervised methods to detect concept drifts. His research is motivated by the challenge of incorporating classification algorithms on embedded devices such as sensors. He has developed a sensor to automatically classify insects in flight using Machine Learning and Data Streams algorithms. This application has led to the development of new algorithms to count events accurately in Data Streams, including the proposal of a novel Data Mining task known as One-class Quantification. He has published more than 100 papers in high-impact venues, and his papers account for more than 6000 citations.