The complexity of the mucus clearance system, with multiple sensors and effectors, requires an integrated model for true understanding. We take as our long range goal the complete integration of our individual models as developed through 8 projects. As defined within this proposal, we take a conservative view of a five year project with the definition of three integrated models as defined in the Research Methods Project 9, with a static model, a cilia flow model and an air-flow model. As with other multi-scale and multi-physics systems, part of the theoretical challenge is to chose the description of the mucosal transport process on a limited set of coarse-grained variables. One then resolves the sub-scale physics and chemistry in some averaged or parameterized form. The manner in which these short timescales and small spatial fluctuations are built into, or suppressed from, the coarse-grained theory and models must be done carefully (i.e., through mathematical upscaling and averaging methods) and tested with new model experiments. We work with a priori recognition that each numerical component must be compatible with future hierarchical linkages. Inevitably, the individual computational models are each run on different computers, written in different languages and have different scale for their inherent time and length scales and computational efficiency. The assembly of a software simulation from cooperating, but independent, component models is known as Model federation. A federated scientific model combines expertise and research products from different scientific domains into a computation system that could not be easily produced by individual researchers from the individual domains. Done manually, this can be a complex and tedious task. For example, researchers at the North Carolina Supercomputing Center have combined five pre-existing environmental models into a federation that produces weather forecasts in real-time. To do this they wrote 80,000 lines of Unix shell script! We will pursue the automation of this manual work facilitated by a software framework that provides reusable modeling abstractions and components. This will provide a framework for continued evolution of the complexity of the individual models while maintaining the interactivity within the integrated model.