Computational intelligence, also known as artificial
intelligence, or AI, is the study of the design of intelligent
agents.
An agent is something that acts in an
environment—such as a mobile robot, a web crawler, an automated
medical diagnosis system, or an autonomous character in a video game.
An intelligent agent is an agent that acts appropriately in
order to satisfy its goals. That is, the agent must perceive its
environment, decide what action to perform, and then carry out the
action.
Perception comes in many modalities—visual, haptic
(touch), speech, textual/ linguistic, etc. Decision making
also comes in many flavors, depending on whether the agent has
complete or partial knowledge of its world, whether it is acting alone
or in collaboration/ competition with other agents, etc. Finally,
taking actions can have different forms, depending on whether the
agent has wheels, arms, or is entirely virtual. An intelligent agent
should also learn to improve its performance over time, as it
repeatedly performs this sense-think-act cycle.
Here in the Laboratory for Computational Intelligence
(LCI), we have been studying all these issues and more since 1981.
Perhaps the most famous contribution of this lab was the invention of
the robotic soccer competition in 1992 (which became Robocup
in 1997). For a more recent synopsis of the
research activities of our 14 faculty members, click here.
|