1 of 22

����A Long-Term Engagement with a Social Robot for Autism Therapy

​

As studied by:

Rakhymbayeva N, Amirova A and Sandygulova A (2021)

2 of 22

Introduction

Definitions

​

Abbreviation

Meaning

ASD

Autism Spectrum Disorder

ADHD

Attention Deficit Hyperactivity Disorder

RAT

Robot Assisted Therapy

ADOS -2

Autistic Disorder Observation Schedule

ANOVA

Analysis Of Variance

NAO

Autonomous programmable humanoid robot

ICC

Intra-class Correlation Coefficient

3 of 22

Introduction

  • Autism Spectrum Disorder is a developmental disability that can cause significant social, communication and behavioral challenges
  • Robo-therapy is an emerging intervention being used to treat behavioral disorders
  • This study focuses on robot assisted therapy for children with Autistic Spectrum Disorder(ASD) and Attention Deficit Hypersensitivity Disorder(ADHD)
  • Longitudinal studies have been proven more efficient than short term studies, however they are not used as often
  • This research uses longitudinal study to investigate effect of robot assisted autism therapy on behaviors of ASD children

4 of 22

Introduction

  • Designing a long-term intervention is complex and the most effective method of investigating human – robot interventions
  • Personalization of the RAT sustains long term user engagement, with engagement score, valence score, unfamiliar and familiar activities used as main indicators
  • Past studies have focused on the technical area of RAT and this study addresses the gap left creating interactive and engaging activities
  • It directs research attention towards individualized approach to fulfil the needs and preferences of children in a socially active and flexible environment with a robot

5 of 22

Methodology

Hypothesis

  • H1: Children will increase their engagement and valence scores when interacting with the robot over multiple sessions
  • H2: Using activities familiar to each child will lead to increased overall engagement from session to session

6 of 22

Methodology

Robot Application Activities

Dances

Songs

Emotions

Touch me

Storytelling

Imitation

7 of 22

Methodology

​

Preliminaries

Therapy Sessions

Video Coding

Measures

8 of 22

Methodology

9 of 22

Methodology

10 of 22

Methodology

Measures

    • Mean of engagement scores per session

Sn Engagement

    • Mean valence scores calculated per session

Sn Valence

    • Amount of time a child is engaged per session relative to overall time

Sn Engagement time

    • Amount of time the child spent looking at the robot relative to overall time

Sn EyeGazeTime

11 of 22

Methodology

Measures

    • An average of all engagement scores

Mean_Engagement

    • An average of all valence scores

Mean_Valence

    • Mean of Sn Engagement Time per child in all sessions

Mean_Engagement Time

    • Average amount of time the child spent looking at the robot throughout sessions

Mean_EyeGazeTime

12 of 22

Results

  • H1: Children will increase their engagement and valence scores when interacting with the robot over multiple sessions
  • Comparison of measurements were performed across sessions for all children
  • Results showed non significance difference in engagement time of the children across all sessions with p value = 0.05
  • Other metrics such as valence score, eye gaze and engagement duration were insignificant with p-value > 0.05

13 of 22

Results

  • H1: Using activities familiar to each child will lead to increase overall engagement from session to session

14 of 22

Result

  • Sn Engagement measurement

​

15 of 22

Result

  • Sn Valence score

​

16 of 22

Result

  • Sn Engagement duration

​

17 of 22

Result

  • Sn Eye Gaze duration

​

18 of 22

Results

Interviews

  • Recorded feedback from 5 parents and the 2 therapists
  • Robot should have cognitive activities, develop more interactive and active behaviors
  • Children should interact with the robot themselves and only observed from outside
  • Incorporate a triadic relationship with an autistic child, non-autistic child and a robot
  • Participants are not to be forced to interact with the robot if they do not want to because of their vulnerable characteristics

19 of 22

Discussion

  • Children were more engaged , positive and focused when we used their preferred activities
  • High engagement levels maybe triggered by children predilections, therefore an autism intervention should be tailored to fit needs and skills of each child
  • The study found no increase in engagement rate of children despite having multi-purpose activities
  • Children preferred familiar activities to unfamiliar ones
  • Autonomy will increase performance of robo-therapy on autistic children

20 of 22

Conclusion

  • Analyze moments when the child is engaged and maximize learning activities at that time
  • Study factors that could increase engagement and take advantage of such variables
  • The study aims to create more active play and cognitive activities
  • It is possible to sustain engagement of an ASD child with multiple sessions of RAT
  • Children are more responsive to activities specific to their needs

21 of 22

Critique

Positive

  • The paper gives a clear background study to address the gap in Robot Assisted Autism Therapy(RAAT)
  • Interesting research topic

​

Negative

  • There were a few grammatical errors
  • Some graphs lacked clarity

​

22 of 22