The analysis of moving objects is a rather challenging but also informative task. It provides important knowledge which is required in a large variety of research fields and applications. The same holds for the analysis of group movements, which, for instance, is able to provide knowledge that can support or contribute to social behavior studies of human beings or animals. In this work, we present an approach to identify changes in the group structure over time, which can support studies regarding an object’s role or standing within the group. The presented methodology is based on the transformation to relative movements and a consecutive change detection algorithm. In order to evaluate our approach, we further conduct experiments by applying it to real data from different scenarios. The results obtained in these experiments demonstrate the functionality and its portability to different uses cases.