One of the common problems for fuzzy system implementation arises from the complications of the fuzzy inference process. Extra computations are required to deduce a consequence due to nature of fuzzy sets. Furthermore, considerable simulations need to be performed to verify system functions. In order to reduce the prototyping time, the fuzzy system is partitioned into hardware and software portions. The model, called Fuzzy Rule-based Automata (FRA), is proposed to simplify fuzzy rule base. Since most rule base are rarely changed, they can be implemented in hardware to speedup the running time. Special computations for inference process is taken care by software to reduce hardware complications which yields prototype flexibility.