Accession Number : ADA189056

Title :   Physics for Robots.

Descriptive Note : Technical rept.,

Corporate Author : BBN LABS INC CAMBRIDGE MA

Personal Author(s) : Schmolze, James G.

Report Date : SEP 1987

Pagination or Media Count : 212

Abstract : Robots that plan to perform everyday tasks need knowledge of everyday physics. Physics For Robots (PFR) is representation of part of everyday physics directed towards this need. It includes general concepts and theories along with an application specific portion for tasks in cooking. PFR goes beyond the representation schemes of most AI planners by including natural processes that the robot can control. It also includes a theory of material composition so robots can identify and reason about physical objects that break apart, come together, mix, or go out of existence. This focus is made on the characterization of knowledge with (1) PFR characterizes the robot's capabilities to act and perceive, and (2) PFR replaces the NP goal of developing models of actual common sense knowledge. The goal of PFR is to represent the knowledge about everyday physics that is needed to prove the effectiveness of certain robot Input/Output programs in accomplishing given tasks. It is expected that this knowledge will be needed by a robot that can plan to perform such everyday tasks. Keywords: Artificial intelligence; Knowledge representation; Reasoning; Planning qualitative physics; Robotics.

Descriptors :   *ARTIFICIAL INTELLIGENCE, *INPUT OUTPUT PROCESSING, *ROBOTICS, REASONING, ROBOTS, DECISION MAKING.

Subject Categories : Bionics
      Cybernetics

Distribution Statement : APPROVED FOR PUBLIC RELEASE