Abstract
The study elaborates on current developments in fuzzy modelling, especially fuzzy rule-based modelling, by positioning them in the general setting of granular computing. This gives rise to granular fuzzy modelling where the models built on a basis of fuzzy models are then conceptually augmented to make them in rapport with experimental data. Two main directions of granular fuzzy modelling dealing with distributed data and collaborative system modelling and transfer knowledge are formulated and the ensuing design strategies are outlined.