This master thesis deals with the issue of tactical capacity planning on the one hand and on the other with robust planning focused on poorly predictable demand. Capacity planning at the tactical level is also known as rough-cut capacity planning. It supports the planer to validate the production plan regarding the limiting capacity factor. The time scope is between six to eighteen months. To do so the available capacity of the production und the capacity demand of the production plan are determined and compared against each other. Due to uncertainty it is not always possible to predict the exact capacity demand. It is necessary to compensate for the negative effects of uncertainty in being stock-out or inventory overflow. Robust planning is one way of compensation by developing a production plan which remains valid for several scenarios. Data aggregation constitutes a focal point of rough-cut capacity planning and robust planning. There a several ways to do data aggregation two of which are presented in closer detail. One is fuzzy set theory, which is well suited to cope with uncertainty in data and covers the issue of robustness. Multivariate methods are widely known and easy to implement models for grouping data in homogenous groups as well as to recognize patterns. Within the scope of the industrial applicability a multivariate aggregation model has been developed. It incorporates a robust methodology to form product families. They should remain unaffected by demand variations. The developed model is applied as an automated computer program on the portfolio of the AMAG rolling inc. The program is part of the company’s rough-cut planning process. It enables the planer to do simulations of different demand scenarios, hence increasing the flexibility, accuracy and robustness of the rough-cut capacity planning process.