Knowledge Processing for Autonomous Robots

Abstract

Autonomous robots are becoming more and more skilled in performing human-scale manipulation tasks, and will soon become common co-workers in our homes. In this work, we describe a framework for representing the knowledge that an autonomous robot needs for performing household tasks, for automated reasoning on this knowledge, and for acquiring new knowledge from sources on the Internet and from observations of human activities. The developed methods allow programmers to write more general and flexible control programs and help robots to execute incomplete and underspecified instructions.