A Novel Reinforcement-Based Paradigm for Children to Teach the Humanoid Kaspar Robot
Author
Zaraki, Abolfazl
Khamassi, Mehdi
Wood, Luke
Lakatos, Gabriella
Tzafestas, Costas
Amirabdollahian, Farshid
Robins, Ben
Dautenhahn, Kerstin
Attention
2299/22194
Abstract
This paper presents a contribution to the active field of robotics research with the aim of supporting the development of social and collaborative skills of children with Autism Spectrum Disorders (ASD). We present a novel experiment where the classical roles are reversed: in this scenario the children are the teachers providing positive or negative reinforcement to the Kaspar robot in order for the robot to learn arbitrary associations between different toy names and the locations where they are positioned. The objective of this work is to develop games which help children with ASD develop collaborative skills and also provide them tangible example to understand that sometimes learning requires several repetitions. To facilitate this game we developed a reinforcement learning algorithm enabling Kaspar to verbally convey its level of uncertainty during the learning process, so as to better inform the children interacting with Kaspar the reasons behind the successes and failures made by the robot. Overall, 30 Typically Developing (TD) children aged between 7 and 8 (19 girls, 11 boys) and 6 children with ASD performed 22 sessions (16 for TD; 6 for ASD) of the experiment in groups, and managed to teach Kaspar all associations in 2 to 7 trials. During the course of study Kaspar only made rare unexpected associations (2 perseverative errors and 1 win-shift, within a total of 272 trials), primarily due to exploratory choices, and eventually reached minimal uncertainty. Thus the robot's behavior was clear and consistent for the children, who all expressed enthusiasm in the experiment.