University of Louisville
Accuracy Analysis of Program Compliance and Additional Developments for Human Robot Interaction in the Observation of Autism Spectrum Disorder
Grade Level at Time of Presentation
Junior
Major
Electrical and Computer Engineering
KY House District #
3
KY Senate District #
7
Faculty Advisor/ Mentor
Dr. Karla Welch
Department
Louisville Automation and Robotics Research Institute
Abstract
Children with Autism Spectrum Disorder (ASD) often have problems with social communication and interaction which can make navigating everyday life challenging. Diagnosing and intervening the disorder as early as possible has a long term positive effect in breaking communication barriers. Thanks to the rapid progression in robotic innovation and research, the use of robots have extended to aiding clinical diagnosis and therapeutic treatment for neurological disorders. Children with ASD also tend to interact better with robots rather than human instructors due to the former's more predictable and consistent behavior. Important components in communication such as eye contact and acknowledgement of verbal and nonverbal cues can become easier to apply in real life human interaction. A recent study was conducted utilizing a socially interactive robot to prompt conversation between children with ASD with the aim to facilitate data acquisition as well as improve social skills, The robot was programmed to detect a slack in the interaction between children by measuring times of silence under certain conditions before bringing up new topics. This work analyses how accurately the robot follows its program to facilitate conversation as well as looks into implementing machine learning algorithms to improve its effectiveness in future studies involving autism, such as motion and gaze tracking to persuade more use of eye contact. The robot has been found to meet expectations when it comes to complying to its current program and a breakthrough in utilizing a face detection algorithm for future head tracking implementation was also successful
Accuracy Analysis of Program Compliance and Additional Developments for Human Robot Interaction in the Observation of Autism Spectrum Disorder
Children with Autism Spectrum Disorder (ASD) often have problems with social communication and interaction which can make navigating everyday life challenging. Diagnosing and intervening the disorder as early as possible has a long term positive effect in breaking communication barriers. Thanks to the rapid progression in robotic innovation and research, the use of robots have extended to aiding clinical diagnosis and therapeutic treatment for neurological disorders. Children with ASD also tend to interact better with robots rather than human instructors due to the former's more predictable and consistent behavior. Important components in communication such as eye contact and acknowledgement of verbal and nonverbal cues can become easier to apply in real life human interaction. A recent study was conducted utilizing a socially interactive robot to prompt conversation between children with ASD with the aim to facilitate data acquisition as well as improve social skills, The robot was programmed to detect a slack in the interaction between children by measuring times of silence under certain conditions before bringing up new topics. This work analyses how accurately the robot follows its program to facilitate conversation as well as looks into implementing machine learning algorithms to improve its effectiveness in future studies involving autism, such as motion and gaze tracking to persuade more use of eye contact. The robot has been found to meet expectations when it comes to complying to its current program and a breakthrough in utilizing a face detection algorithm for future head tracking implementation was also successful