cell system, whose behaviour is characterized by a number of macro events
(e.g. WORKDONE ) that trigger transitions between macro states (e.g.
WORKING and NOT WORKING ). The work cell simulator does not animate the
execution of manufacturing operations but simply represents their start and
completion times. The behaviour of the robot system, however, is character-
ized by a few macro states (e.g. PLATFORM ARRIVED ) and a large number of
micro states (e.g. every single rotation step).
How do we model the micro and macro behaviours of the robot system?
We can generalize the macro behaviour of a robot's device with the finite
state machine described in Figure 10.13 (left-hand side). This is character-
ized by two states, RUNNING (device is executing a task) and NOT RUNNING
(the device has completed the last task and there are no more tasks to
execute), and two events, NEW TASK ( NT ) and TASK COMPLETED ( TC ). As soon
as a new command is issued it is executed immediately if the device is not
running, or it is inserted in a command list if the device is running. As soon as
a task gets completed a new one is started unless the command list is empty.
When the device is running the simulator animates the micro steps that
discretize the task execution. The number of steps depends on the device
precision. For example, if the platform has a resolution of one degree, the
complete rotation around the vertical axis is simulated with 360 micro
rotation steps. The duration of each micro step depends on the device speed.
We model the execution of a task with the finite state machine depicted in
Figure 10.13 (right-hand side). It has only the TICK event and three states:
the INITIAL state, the FINAL state and the WAITING state. The device is in the
waiting state for the entire duration of a micro step, which is executed when
the tick event is raised.
Which execution model is best suited to simulate the robot system?
commands p 0
steps p 0
commands # 0
steps # 0
Figure 10.13 The finite state machine describing a device's behaviour