One underlying idea of this specification is that a vision system may have different modes of operation with very different sets of states, methods and transitions.

Due to the high degree of individuality of vision system requirements and solutions, it does not appear possible to standardize each and every mode of operation of such systems. Therefore, vision systems according to this specification are free to implement additional state machines for such use cases.

However, this specification considers some states as universal for machine vision systems according to this specification and thus outside of these modes of operation, namely states related to powering up and shutting down the system and to error handling.

The system will in any case need to power up and enter some state automatically without outside intervention. From this state, a selection – either by method call or automatically based on internal and external circumstances – of the actual mode of operation can be made.

Conversely, the same holds for shutting down the system. Independent from the mode of operation, the system will need a way to enter a state from which it can safely be powered down.

And finally it will be of great advantage to all clients if the handling of errors is identical in all modes of operation.

Therefore, these states are mandatory in this specification as well as one state encompassing the actual operation of the system. Modes of operation shall be specified as SubStateMachines to this operational state.