Novel products are commonly realized by integrating heterogeneous technologies. Product architecting focus on defining the scheme by which the product functions are allocated to physical components. A DSM-based clustering method that integrates technical complexity and strategic concerns has previously been proposed. It has been shown that interaction weights in the DSM may affect the clustering result. A complexity-based interaction strength model to be used in DSM clustering is proposed here. The case study gives promising results from both interaction performance and safety points of view.