ADAPTIVE ROBOTICS FOR CHILDREN WITH AUTISM

Overview

The techniques for developing behavioral models from physiological signals and applying those models to recognize an individual’s behavioral states have provided highly reliable results in previous work for typical adults (Rani et al., 2006). Questions remain regarding if such techniques could prove reliable for a younger population or a population with hindrances understanding or expressing such behavioral states. A pilot study was conducted with children with Autism Spectrum Disorders (ASD) to develop affective models based on their physiological signals, which produced highly reliable results as well (Liu, Conn, Sarkar, & Stone, 2007). Social communication and social information processing are thought to represent core domains of impairment in children with ASD. There is a need to better understand the underlying mechanisms and processes associated with these deficits as well as develop tools that can be used to create optimal intervention strategies. Our current endeavors utilize and merge recent technological advances in the areas of (i) virtual reality, (ii) physiological signal processing, (iii) machine learning techniques, and (iv) adaptive response technology in an attempt to create a tool for understanding various physiological aspects of social communication in children with ASD.

Related publications:

• Liu, C., Conn, K., Sarkar, N., & Stone, W., "Affect Recognition in Robot Assisted Rehabilitation of Children with Autism Spectrum Disorder," In Proc. IEEE International Conference on Robotics and Automation (ICRA 2007). April 2007, Roma, Italy, pp. 1755-1760.
• Liu, C., Conn, K., Sarkar, N., & Stone, W., “Online Affect Detection and Adaptation in Robot Assisted Rehabilitation for Children with Autism,” in Proc. 16^th IEEE International Symposium on Robot and Human Interactive Communication (RO-MAN 2007), Jeju Island, Korea, August 2007.

Top

Mechanical Engineering Home | School of Engineering Home
Vanderbilt University Home