fcrarproceedingstemplate-word.docVIP

Artificial Intelligence Grounded in Reality A. Antonio Arroyo Director, Machine Intelligence Laboratory University of Florida 00-352-392-2639 arroyo@ Eric M. Schwartz Associate Director, MIL University of Florida 00-352-392-2541 ems@ ABSTRACT At the Machine Intelligence Laboratory (MIL) in the Department of Electrical and Computer Engineering at the University of Florida (UF), we seek to develop fundamental knowledge and gain practical experience in the design, realization, and application of intelligent, autonomous, sensor-driven, behavior-based robotic agents. These agents generate knowledge, share information and learn to improve their performance with experience. We seek to combine a wide variety of advanced technologies to give computer-driven autonomous machines the ability to learn, adapt, make decisions, and display behaviors not explicitly programmed into their original capabilities. We use the term agent and robot interchangeably. Autonomous refers to the independent control of each agent/robot by its embedded computer. (The term ‘robot’ comes from a Czechoslovakian word meaning “forced servant,” and was first coined in the play Rossum’s Universal Robots by the writer Karel Capek in 1920). The term ‘mobile’ refers to the ability of the robots to maneuver around an environment un-tethered and self-reliant. In this context we define machine intelligence as the ability to build machines that exhibit a high degree of sophistication and can operate autonomously in ‘their’ environment. For example, a roach is a highly intelligent insect (as evidenced by the US$100 million industry designed to attempt to control them, especially in Florida), and thus, a ‘roach-like’ machine would also be considered ‘intelligent.’ At MIL, we define artificial/machine intelligence is simply “computer intelligence grounded in reality,” which answers the question ‘what intelligence can be realized?’ Keywords Robot, machine intelligence artificial intelligence, autonomous. THE M