||Most of the existing tools for modeling manufacturing systems are based on operations research and control theory, including the Markovian model, the network model, and perturbation analysis; in the past several years, the Petrinet model has also been used increasingly. While all these models can facilitate quantitative analysis, they do not capture the qualitative behavior of the systems. Further-more, these models are too fEnt to be effective in handling manufacturing systems with very large number of cells, objects, and states. Moreover, none of the above models supports dynamic expansion of the system adequately, and they fall short of providing a clear conceptual view/picture of the entire system as it grows. In order to complement these models, a more generic framework, called the Extended Cellular Automaton Model (ECAM), for manufacturing system modeling is proposed and investigated in this thesis. ECAM is based on the merits of the cellular automaton theory with extensions on the enhancement of its expressive power to capture the concurrency and dynamic nature of manufacturing systems. In this thesis, ECAM will be defined both mathematically and visually. On top of ECAM, an object-oriented extension is proposed to provide a more natural methodology to specify extremely large and complex systems, as well as for conceptual modeling. The visual ECAM model is used as a visual specification langauge for model creation. Based on this model, the resolution schemes for concurrency problems are discussed in details. Model execution is facilitated and based on the translation scheme in which ECAM will be automatically translated to a corresponding set of rules, such that these rules can be run on any rule-based system. This execution scheme is so clean a.nd general that rapid prototyping, simulation, validation, verification, and analysis can be easily achieved by using rule-based or expert system techniques. Finally, some autonomous assembly systems are used as examples to illustrate the applications of ECAM.