This paper addresses a common problem to manufacturing companies: the maintenance of machine tools and their components. Preventive maintenance has always been a great challenge for companies, due to the need of predicting failures or production shutdowns, which requires knowledge and resources. However, the planning of machine tools maintenance presents itself as an even more complex problem due to the distinct lifetimes of their components. Age-based preventive replacement and Block replacement models define optimal replacement intervals for one item based on associated maintenance costs. A machine tool can be seen as a serial system of components or items. The concepts of group technology and clustering can be used to group components together in order to define common preventive maintenance intervals and reduce the number of production stops. In the literature, some contributions are found. However, the defined groups are static as well as the preventive maintenance intervals. This paper presents a conceptual model for the definition of dynamic clusters and intervals. It also presents an application to record the inputs, data collected in real time, needed to group components and set up preventive maintenance intervals. The developed application is being implemented in a metalworking company.