The paper presents contributions in the area of location data processing for pattern discovery. This work forms part of a project which explores an ambient intelligence application designed to present individual users with an overview of their time usage patterns. The application uses location data to build interfaces and visualizations which highlight changes in personal routines, with the aim of stimulating reflection. Data is processed to extract significant places and temporal information about them. The paper presents the questions that can be answered by a data processing layer and the strategy to handle the different types of queries. Location data is processed to identify significant locations, discover patterns and predict future behavior.