![]() |
The following pictures show examples of mobile intelligent agents interacting with each other in different distributed scenarios. The agents are free to roam between computers running the Wave interpreter. For demonstration purposes only two computers have been used to represent the scenario space (which is stored as a grid of locations), although any number of machines (up to the number of positions in the space) can be configured. The vertical black line in the diagrams denotes the boundary between the computers.Each software agent may have arbitrarily complex behaviour. In these demonstrations, only simple behavioural rules are used for the individual agents, but collectively they behave as an intelligent group.
It must be remembered that in these examples, each "agent" is a self-contained entity. There is no one program controlling all their behaviour.
A group of ten agents are deployed to dig a ditch. They cooperate to ensure all the "soil" is removed from the specified area, then line up in a finishing position. This kind of behaviour may be useful for remote controlled robots.
A group of agents have to move from one side of the space to the other. However, a maze of arbitrary complexity lies in their path. They must negotiate the maze, cooperating in order to avoid deadlocks.
A group of lightweight agents (dismounted infantry) is ordered to de-aggregate, move to a new location, and then re-aggregate. A group of heavyweight agents (column of tanks) is crossing their path. The lightweight agents are more agile, and infiltrate the column, breaking up its order.