Agents & Robots

Agent: An agent is a computer system that is situated in some environment, and that is capable of autonomous action in that environment in order to meet its delegated objectives.

Agents can make their own decisions. Tele-operated systems aren’t agents.

Mobile Robotics

There are three main questions in mobile robotics:

Agents are on a scale of autonomy:

Agent	Object
Agents are autonomous.	Objects do it because they have to.
Agents are smart (proactive, reactive, social).	Objects only act on request.
Agents act because they get a reward.

Intelligent agents exhibit three types of behaviour:

This is the act of generating and attempting to achieve goals, not solely by events.

This also includes recognising opportunities to achieve it’s goals.

If a program’s environment is fixed then it can act deterministically. We can react to changes in the environment.

This is the ability in agents to interact with other agents via cooperation, coordination and negotiation.

The basis of this is communication.

Cooperation: Working together as a team to achieve a goal.
Coordination: Managing the interdependences between activities (what should be done in what order).
Negotiation: The ability to reach agreements on matters of common interest.
Typically involves offer and counter-offer, with compromises made by participants.

Mobility: The ability for agents to move around a non-deterministic environment.
Rationality: Whether an agent will act in order to achieve it’s goals.
Veracity: Whether an agent will knowingly communicate false information.
Benevolence: Whether agents have conflicting goals or are working together.
Learning/Adaptation: Whether agents improve performance over time.