Proper orthogonal decomposition (POD) is an attractive way to obtain nonlinear low-dimensional models. This article reports on the automatization of the mentioned reduction method. An automatic procedure for the reduction of differential algebraic systems is presented, which is implemented in the modeling and simulation environment ProMoT/Diana. The software tool has been applied to a nonlinear heat conduction model and a continuous fluidized bed crystallizer model. The automatically generated reduced models are significantly smaller than the reference models, while the loss of accuracy is negligible. (C) 2016 Elsevier Ltd. All rights reserved.