It is well accepted that modelling of complex systems involves transition from detailed numerical to more abstract, aggregated information processing. Granular computing is geared toward representing and processing aggregated chunks of information - information granules. These granules are collections of entities that are arranged together due to their similarity, functional adjacency, indistinguishability or alike. No matter how the information granulation proceeds and what fundamental technology becomes involved therein it is driven by an overreaching need to split the problem into more manageable subtasks and to provide a better insight into its nature. The rapid development of granular computing in the late 90's arises from the fact that unlike numeric computing, which is data-oriented, granular computing is knowledge-oriented, and as such, it underlies knowledge-inclined system modeling. I envisage that with the increased visibility of granular computing, granular models are about to come to existence in a broad spectrum of application domains. Consequently, development of a well-rounded design environment represents a challenging yet highly rewarding long-term target. There are many fundamental questions that will be asked: Are granular models preferred over some other approaches? How should a hierarchy of information granules be organized? How to validate the models against experimental and perception-driven evidence? What is the best means of communication between granular models? The motivation of this research is to make such investigation more structured and comprehensive. The detailed objectives that can be set in this research program are: