Report
Teaser, summary, work performed and final results
Periodic Reporting for period 1 - CARER-AID (Controlled Autonomous Robot for Early detection and Rehabilitation of Autism and Intellectual Disability)
Teaser
This project experimented a humanoid robot as a supervised autonomous assistant to support caregivers in early diagnosis and to improve the treatment of individuals with Autism Spectrum Disorder (ASD) associated with Intellectual Disability (ID). The robot introduced as part...
Summary
This project experimented a humanoid robot as a supervised autonomous assistant to support caregivers in early diagnosis and to improve the treatment of individuals with Autism Spectrum Disorder (ASD) associated with Intellectual Disability (ID). The robot introduced as part of the diagnostic team during the administration of the psycho-diagnostic tests in order to enrich the data that the psychologist can use to refine the diagnosis, helping them to perform early diagnosis and to distinguish among the ASD types and ID levels. This research program was undertaken by the Marie Skłodowska-Curie Experienced Researcher under the supervision of a multidisciplinary support team composed by the academic staff of the Sheffield Robotics in partnership with IRCSS Maria SS Oasi of Troina (Italy). The highly interdisciplinary program blended the consolidated international experience of Sheffield Robotics in the responsible application of robotics and human-robot interaction with the Researcher’s experience in the theory and practice of psychological assessment, patient rehabilitation and use of robots in the education and therapeutic care of individuals with ASD and ID. The final output of the research project was a set of use cases that will be implemented and empirically validated via pilot studies and proof-of-concept trials in kindergarten and clinical environments. Furthermore, the project accomplished a series of actions for the widespread scientific dissemination of the experimental results and outreach activities to give evidence also to the general public, especially targeting families with individuals affected by ASD and ID, of the actual opportunities offered by robot and, thus, increase their acceptance and willingness to use robots in the care of children and disabled.
Work performed
\"The work has mainly focused on two sectors highly connected to each other: teachers and students acceptance and attitude of the robots, and children with typical developmental or with ASD and ID.
At the beginning the aim was to evaluate the predictive factors and attitudes of curricular and specialized teachers towards socially assistive robotics and the intention to use robots in teaching activities. The results highlight the primary role of the personality factors Openness to Experience and Extraversion for promoting the acceptability and reduce the prejudicial reject regarding the use of educational and assistive robotic technologies.
Subsequently we have moved our interest on children, before with typical development and after in children with ASD and ID.
We found that children exposed to the robot decrease their distress and positively change their attitude toward the technological device. This suggests that an early, controlled exposure may facilitate future acceptance. In another study we presented an experimental study with 81 kindergarten children on memorizations of two tales narrated by a humanoid robot. Results suggest a positive effect of the expressive behaviour in robot storytelling, whose effectiveness is comparable to a human with the same behaviour and better when compared with a static inexpressive human. Higher efficacy is achieved by the robot in the tale with knowledge content, while the limited capability to express emotions made the robot less effective in the tale with emotional content.
To support the therapists, we integrated a humanoid robot within the standard clinical treatment of children with ASD. In this research, using the A-B-A1 single case design, we proposed a robot assisted affect recognition training and present the results on the child progress during the 5 months of the clinical experimentation. The results of the single case study suggest the feasibility and effectiveness of using a humanoid robot for emotion recognition training in children with ASD.
Subsequently we focused on children with ASD and Intellectual Disability (ID). The experiment integrated a robot-assisted imitation training in the standard treatment of six hospitalised children with various level of ID, who were engaged by a robot on imitative tasks and their progress assessed via a quantitative psychodiagnostic tool. Results show success in the training and encourage the use of a robotic assistant in the care of children with ASD and ID with the exception of those with profound ID, who may need a different approach.
Furthermore, we investigated on the use of novel deep learning neural network architectures for automatically estimating if the child is focusing their visual attention on the robot during a therapy session, which is an indicator of their engagement. To study the application, the authors gathered data from a clinical experiment in an unconstrained setting, which provided low-resolution videos recorded by the robot camera during the child–robot interaction. The proposed approaches demonstrated a very high accuracy and it can be used for off-line continuous assessment during the therapy or for autonomously adapting the intervention in future robots with better computational capabilities.
Presentation at Scientific Conferences
ACM/IEEE International Conference on HRI, Daegu-South Korea, 3/2019.
19th Towards Autonomous Robotic Systems (TAROS), Bristol-United Kingdom, 7/2018.
Workshop on Intelligent Assistive Computing, Rio De Janeiro –Brazil, 7/2018.
ACM/IEEE International Conference on HRI, Chicago-USA, 3/2018.
ACM/IEEE International Conference on Human-Robot Interaction, Vienna-Austria, 3/2017.
Participation at the \"\"European Researchers\' Night\"\" event @ IRCSS Oasi of Troina - September 2018
Organization of 3rd Workshop on Behavior Adaptation, Interaction and Learning for Assistive Robotics (BAILAR) - Co-located with the International conference Ro-Man 2018 (Nanjing, China, 27-30 August)
Invited Talk:\"
Final results
The work in this project contributes to alleviating the increasing concern worldwide about the diagnosis and treatment of children with ASD by proposing the use of robot assistants.
Toward its vision, this project introduced three main innovations: (i) designed a set of use cases in which a socially assistive robot gives support to the diagnosis and rehabilitation of ASD and ID; (ii) implemented novel control strategies for autonomous and safe robot-child interaction that can support the intelligent personalization of activities; (iii) validated the use cases with the integration of the robot in everyday activities and standard therapeutic protocols in a healthcare center.
This innovative project offered a unique vision in which the robot can serve in all the levels of the care, from the diagnosis to the therapeutic rehabilitation, and provide assistance to caregivers at school, clinic, and home. This was an original theme that constitutes an advancement over most of the other research programmes that usually focus only on the treatment.
The technological advancement proposed by the research programme is the introduction of autonomous behaviours that will improve usability and overcome the limitations of the WoZ technique. Moreover, the proposed research explicitly addresses patients affected by ID, which lead to significant differences in cognitive and adaptive skills respect to individuals with ASD only.