In a large group, many behaviors are specified in terms of interactions instead of single agent goals.

Steer evolution of desirable traits. The agents can cover the search space of desirable behaviors faster if they do different things.

Reinforcement learning.

Different response thresholds. Allows for a smooth behavior transition in a team setting. For example, as the problem grows in size, more agents are automatically recruited (due to having varying response levels).

In some tasks, variability is desired or required to accomplish it. Having individual variability saves the operator from specifying specific behaviors to each robot. They can self organize to complete a high level task.

Recruitment. Can get n agents to do a task without explicitly assigning n agents. Vital for scalability of the system from an operational perspective.

Allows control of large groups without controlling large numbers of individuals.

Allows operator to shape the behavior of the group. Reinforcement.

Agents automatically reinforce positive behaviors. Allows operator to specify high level behaviors (such as laying trails) that agents then use.

Hysteresis effects. Allows to group to “latch onto” certain behaviors and resist perturbations of the system. Allows groups of agents to reach a steady state, allowing the operator to go focus on something else. Allows for memory in the system. Allows large groups to lock into certain behaviors.

Operators can manipulate thresholds, increasing or decreasing the likelihood that a large global behavior will be adopted.

Allows information to quickly propagate through a group. “Because you're doing it, I'm going to do it to” Allows information to pass between agents even though the operator hasn't (or can't) send the information directly.

Negatively reinforce undesirable behaviors to shape group behavior

Decrease the effect of perturbations of the system.

Allows operators to hold the group in a certain state.

Reverse the effect of a previous Hysteresis change. Tell the group to “look for new options”

Stabilizes the group by requiring some consensus before a certain behavior is adopted.

Operator can directly (or indirectly) modify thresholds the perceived environment of the agents increasing or decreasing the likelihood that a certain behavior will be adopted.

Agent can notice that it's not behaving like the rest of the group and comply.

Provides a way of specifying the importance of the task.

Provides a way to to specify how many agents should respond (individual thresholds vary from agent to agent).

Information propagation mechanism. Once consensus is reached, information propagates quickly.

One agent (the operator or another agent) must be able to communicate information and actions to the rest of the group. Agents must not go with the flow all the time.

Allows for agent autonomy. Agents can try out new ideas. Good ideas recruit others and pass information throughout the group. This decreases the overhead for the operator.

Simplifies control problem for operator. Operator interacts (directly or by means listed above) with small group of leadership prone agents that then propagate information or recruit other agents.

Stabilizes positive feedback. Allows agents to split tasks among group and transition from one task to another. Positive feedback alone would cause the whole group to quickly change behaviors. Leadership within groups provides a stabilizing force that allows the group to break into subgroups.

Provides group coordination. Allows an agent to go “fetch” others and bring them to the group.

Another information propagation mechanism. Allows the operator to say “don't do this”, or “don't do this as much” or “this area is off limits”

Informed agents or leaders (or operators) can steer the group away from undesirable behaviors and towards desirable ones. “don't go that way, go this way”

Provides another stabilizer for positive feedback.

Provides group stabilization. Positive feedback inhibits deviance from the current behavior.

Decreases operator burden by ensuring a task will be carried out, even if the original agent responsible is unable to do so.

Makes the group more resilient to disturbances. Stabilizes behaviors. If the leader that originally initiated the action is knocked out, the group will still continue the behavior (assuming the system is in a stable state). The next most leadership prone individual may take charge.

Allows an operator to pick certain agents to perform tasks. Taking away one agent will not change the global behavior of the group. Allows the operator to focus on one agent without worrying about the rest of the group.

Synchronizing large numbers of individual elements is hard to do (try to synchronize 20 pendulums by yourself :)). Automatic synchronization of tasks frees the operator from micromanagement.

Group forming - allows operator to command a group instead of individual units, simplifies operator control problem.

Reduces need to closely monitor agents in critical situation. For example, coordinated surveillance or attack

Agents can synchronize with operator controlled agent or operator actions. “do this”

Operator can establish patterns and routines that the agents follow.

Provides a reward structure. Operator can modify agents using rewards and punishments.

evolution models

Provides a framework for individual initiative. Brings in utility theory.