Situative Space Tracking within Smart Environments

Abstract

This paper describes our efforts in modeling and tracking a human agent's situation based on their possibilities to perceive and act upon objects (both physical and virtual) within smart environments. A Situative Space Model is proposed. WLAN signal-strength-based situative space tracking system that positions objects within individual situative spaces (without tracking their absolute positions) distributed across multiple modalities like vision, audio, and touch is presented. As a proof-of-concept, a preliminary evaluation of the tracking system was performed by two subjects within a living-laboratory smart home environment where a global tracking precision of 83.4% and a recall of 88.6% were obtained.