At least as important as its contribution to the field of neural nets, is the influence CMAC has had in providing the conceptual foundation for the Real-time Control System architecture known as RCS. The fact that CMAC could learn a set of transfer functions and transition matrices meant that a CMAC could be designed to implement any state-table or expert system rule base The CMAC input vector (command, feedback, and state) corresponds to the IF predicate. The output vector corresponds to the THEN consequent and next state. At each compute cycle, the CMAC input vector is compared with the lines on the left side of a state table. For the matching line, the right side of the state table provides the output subcommand and next state. Thus, any CMAC can be emulated by a finite state machine. RCS is a control system built from a hierarchy of finite state machine modules.