Sep 20 2010
Researchers, Wei-Po Lee, Tsung-Hsien Yang and Bingchiang Jeng at Taiwan’s National Sun Yat-sen University, have created a user-based communicative framework that utilizes a neural network-based method in robot toys, eliminating its repetitive behavior and offering interactive responses to human instructions.
The research team had tested their robotic structure by developing robot controllers to perform diversified actions. They are now researching to develop a robot that can identify human facial reactions and react accordingly. It was expected that such an innovation will have a major impact on future communicative devices and functional robots.
The major challenge in developing this design is the long duration taken in contrast to the short life cycle of the given product. Other drawbacks include its preset archetype and monotonous performance.
In the recent issue of the International Journal of Modelling, Identification and Control, the researchers had explained their research work and it was expected that their initiatives will lead to the development of a social robot which reacts more quickly than the present design.
Wei-Po Lee and his fellow researchers said that personified archetypes of man-made living robotic toys can be developed by utilizing the latest computing and electronics technologies. The team added that the important issues involved in robot designing include developing the optimum control architecture providing logical sense to the robot, creating innate methods for effective robotic interaction, incorporating sensitive reactions and conduct into the robot’s software. They are likely to deal with all these concerns by implementing behavior-based structural design, which enables them to customize the robot to develop new manners and assuring instantaneous operations in real-time situations.
Source: http://www.oia.nsysu.edu.tw/english/