The CaseLP Visualizer [17]
provides documentation about events that happen at the agent
level during MAS execution.
According to the developer needs, the code of some
agents can be automatically instrumented after it has been loaded.
Instrumentation adds probes to agents code; events related to
state changes and/or exchanged messages can be recorded and collected for
on-line and/or off-line visualization.
Windows for initializing some agents mail-boxes and for setting the
simulation length and the on-line visualization granularity
are
provided.
During the simulation, views related to
instrumented agents are shown. At the end of the simulation a more complete
trace of all the instrumented events can be visualized. Instrumentation
is completely independent from execution. This will not influence a
possible future change of the execution support.
Figure 2 shows the off-line trace of execution of the prototype for
train traffic management [4].
Figure 2: Off-line trace of execution.