Time and Space Reasoning for Ambient Systems

Abstract

This paper presents an algebraic language, called Time-AgLOTOS, to describe time-dependent behavior of intelligent agent for the design of Ambient Intelligence systems. This specification model provides a theoretical foundation for performing planning under timing constraints. Based on a true-concurrency semantics, a contextual model, called Spatio-Temporal Planning System (STPS), is developed to capture all possible evolutions of an agent plan including context changes. The STPS provides formal description of possible actions to perform supporting timing constraints, action duration and spatial information. This structure offers new possibilities and strategies for taking agent real-time decisions in context-awareness manner.