for a Home Robot

• Main Authors: Andriy Bondarev, Ko Crucq, Loe L. M. G. Feijs, Christoph Bartneck
• Other Authors: The Pennsylvania State University CiteSeerX Archives
• Format: Text
• Language: English
• Subjects: Artificial Intelligence; Emotions; OCC Model; Emotional State; Robot; Facial Expressions Abstract; DML
• Online Access: http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.145.7656; http://www.bartneck.de/publications/2002/designEmotionManagementSystem/Bondarev_Final_Report.pdf

This document describes the General Emotion Management System (GEMS) that was created for use in the "Mobile Intelligent (Inter)Face " project. This is the softwarebased prototype of a general emotion management system for a robot that includes Artificial Intelligence (AI). GEMS has been created as an independent software module in the robot's architecture using the Dynamic Module Library (DML). The GEMS module generates an emotional state of the robot each moment of actual time. The emotion generation algorithm is based on the work of O’Reily and on the OCC (Ortony, Clore & Collins) model. As input the GEMS module receives information about the current robot state (goals, actions, events, objects or agents) from the superior robot system and detailed information from a database, which must contain all necessary variables to calculate the intensity value for each emotion. The emotional state of the robot consists of 22 types of emotion in a multi goal environment. The GEMS module produces output in two views: the first view contains information about the emotional state in general and the second view contains information about the dominant emotion. The output of the GEMS module can be