Digital models have replaced the architectural drawing as dominant means of representation and creation. Computation provides models the ability to loop, sense and query and the integration of near real-time simulation provide these models with a depth and agility that allows for instant and informed feedback. Models appear to be almost alive. However models need to be informed and finely calibrated in order to be aware of the tasks and design parameters in question. Based in the practice of CITA this PhD investigates the underlying concepts for the implementation and modelling of material aware models. A focus is set on design processes and computational techniques, such as machine learning, that allow to model novel material systems and highly interrelated models.