Design of Robot Head for Expression of Human Emotion

Autor: Goh Wei Jun, Fariz Ali
Rok vydání: 2015
Předmět:
Zdroj: Applied Mechanics and Materials. :893-897
ISSN: 1662-7482
DOI: 10.4028/www.scientific.net/amm.773-774.893
Popis: Humanoid robot is a type of robot which designed in human-form with the purpose to increase the quality of human life. The key features of humanoid robot are to perform human-like behaviours and to undergo effective interaction with human-operator. Facial expressions play an important role in natural human-robot communication as human communication in daily life relies on face-to-face communication. The purpose of this study was to develop an interactive robot head that able to express six basic human emotions based on Ekman’s model which are joy, sadness, anger, disgust, surprise and fear. The combination of action units based on different control point on robot head was proposed in this study. The new robot head provided with 11-DoFs to perform different expression in human-like way. A survey was conducted on twelve sets of emotion design drawn by using Solidworks. Evaluation had been done on each design for its expression ability and the best design of emotion to implement on the robot head was obtained in the end of survey. Hardware experiment was conducted to control the LCD display and position of servo motor by using Arduino Leonardo as the controller for the robot head system. Additionally, a keypad controller was designed to control the expression of robot head based on the control from user. The controller is connected with LCD display to show the name of facial expression for the learning purpose of autism children. This project focus on the performance test of robot head in term of position accuracy for the 11 actuators used in robot head construction. The result shows that the relative position error for each robot head parts was less than 10% and thus robot head able to perform the emotion effectively. The survey on the recognition rate for each emotion expression was conducted individually to 100 respondents. The recognition rate obtained for the six emotions express by robot head was more than 70%.
Databáze: OpenAIRE