Emotional memories in autonomous robots

Abstract

This is a theoretical study aiming at proposing a computation model for memory and emotion, using ideas inspired by neuroscience research into neural-endocrine systems interaction. Human robot interaction (HRI) ideally requires the robot to be able to detect emotional changes and understand the emotional implications of these changes. The model proposed here will allow a robot to create its own emotional memory, and to use that memory to predict future emotional states based on past experiences. This model would be able to monitor the emotional state of a person, to identify if that individual is in a state of flow, providing positive support and at times praise that may be needed. Further evaluation and validation are proposed including a study case.