ANML Documentation

Character Recognition in an STE

In the world of automata, an STE can be programmed with only one distinct character set. This has subtle implications when working with complex state machines. For example, consider a traditional state machine with three states labeled 1, 2, and 3. State 1 transitions to state 3 with the input A, and state 2 transitions to state 3 with the input B:

Figure 1. STE Character Recognition
Automata character recognition 1

When a fourth state and associated transition is added, state 1 is split into two STEs (one that responds to the input character B and activates the STE corresponding to state 4, and one that responds to the input character A and activates the STE corresponding to state 3):

Figure 2. STE Character Recognition
Automata character recognition 2

State 1 now transitions to state 4 on input B and to state 3 on input A. Any activation connections to STE 1 in the original design now need to connect to both 1a and 1b in the automata design.

There are times when modifications like this will need to occur when converting a traditional state machine into an automaton. The easier and recommended path is to construct the original design in automaton form. It is possible to convert a complex state machine into an automaton, but it may not be easy.