Controlling a Social Robot - Performing Nonverbal Communication through Facial Expressions

Abstract

This paper focuses on our primary goal to achieve the emotional behaviour of the new version of the social robot Probo. The ability to enhance nonverbal communication with children is possible through facial expressions, eye-tracking and face-to-face contact. The new social robot has 21 degrees of freedom (DOF), grouped in five subsystems, named generically: eyes, ears, trunk, mouth and neck. The robotic head is actuated using only servo motors and all the components are manufactured using cheap, flexible and easy technologies. In order to get the social robot head able to express emotions, a Graphical User Interface (GUI) was developed. In this way facial expressions are created through sliders or push buttons. Additionally, we investigated the possibility of controlling the robot with an Arduino board. In this case, using pre-programmed or learned algorithms, the robot is getting a semiautonomous level, based on the usage of various sensors, being able to express six basic emotions: happiness, sadness, fear, anger, surprise and disgust. So, based on the feedback provided by the sensors, the robot can react accordingly, enhancing human-robot interaction (HRI).