Long-Term Personalization of an In-Home Socially Assistive Robot for Children With Autism Spectrum Disorders

Abstract

Socially assistive robots (SAR) have shown great potential to augment the social and educational development of children with autism spectrum disorders (ASD). As SAR continues to substantiate itself as an effective enhancement to human intervention, researchers have sought to study its longitudinal impacts in real-world environments, including the home. Computational personalization stands out as a central computational challenge as it is necessary to enable SAR systems to adapt to each child's unique and changing needs. Toward that end, we formalized personalization as a hierarchical human robot learning framework (hHRL) consisting of five controllers (disclosure, promise, instruction, feedback, and inquiry) mediated by a meta-controller that utilized reinforcement learning to personalize instruction challenge levels and robot feedback based on each user's unique learning patterns. We instantiated and evaluated the approach in a study with 17 children with ASD, aged 3 to 7 years old, over month-long interventions in their homes. Our findings demonstrate that the fully autonomous SAR system was able to personalize its instruction and feedback over time to each child's proficiency. As a result, every child participant showed improvements in targeted skills and long-term retention of intervention content. Moreover, 1 arXiv:1911.07992v1 [cs.RO] 18 Nov 2019 Clabaugh et al. Long-Term Personalization of a Home Robot for Children with ASDall child users were engaged for a majority of the intervention, and their families reported the SAR system to be useful and adaptable. In summary, our results show that autonomous, personalized SAR interventions are both feasible and effective in providing long-term in-home developmental support for children with diverse learning needs.

Collaborative Research Project: Developing and Testing a Robot-Assisted Intervention for Children With Autism

“…In further studies, robots should be endowed with an extensive set of educational goals and sensory options so that the administration of the educational procedure can be personalized. A first step toward this goal was recently made by Clabaugh et al ( 2019 ) who developed a fully autonomous robot, SPRITE, able to personalize its instruction and feedback to each child's proficiency.…”

“…To date, nearly 30 robots were tested as remedial tools for ASD [e.g., : Labo-1 (Werry et al, 2001 ); Muu (Miyamoto et al, 2005 ), Robota (Billard et al, 2007 ), FACE (Pioggia et al, 2007 ), Keepon (Kozima et al, 2007 ), Aibo (Francois et al, 2009 ), IROMEC (Iacono et al, 2011 ), Charlie (Boccanfuso and O'Kane, 2011 ), NAO (Shamsuddin et al, 2012 ), Flobi (Damm et al, 2013 ); GIPY-1 (Giannopulu, 2013 ), Pleo (Kim et al, 2013 ), KASPAR (Wainer et al, 2014 ), Darwin-OP (Peng et al, 2014 ), Pabi (Dickstein-Fischer and Fischer, 2014 ), Zeno (Salvador et al, 2015 ), Jibo (Guizzo, 2015 ), Probo (Simut et al, 2016 ), Maria (Valadao et al, 2016 ), Sphero (Golestan et al, 2017 ), CARO (Yun et al, 2017 ), KiliRo (Bharatharaj et al, 2018 ), MINA (Ghorbandaei Pour et al, 2018 ), QTrobot (Costa et al, 2018 ), Milo (Chalmers, 2018 ), Leo (She et al, 2018 ), Daisy (Pliasa and Fachantidis, 2019 ), SAM (Lebersfeld et al, 2019 ), SPRITE (Clabaugh et al, 2019 ), Actroid-F (Yoshikawa et al, 2019 ) etc.].…”

AI and the Future of Learning: Expert Panel Report

“…In a third possible definition, AI also empowers a "science of the artificial" (Simon, 1969(Simon, /1996 where innovators can create new learning environment configurations in order to study what the future could bring. Some examples include the possibility of students learning collaboratively with an artificial agent that facilitates their social interactions, for example, preschool-age children learning science with a social robot who motivates them and supports their inquiries (Kim et al, 2018), or differently-abled learners getting personalized support from a near-peer socially assistive robot buddy (Clabaugh et al, 2019). AI can also be a toolkit for building innovative approaches to assess students' competencies (Paquette, et al, 2014;Mislevy, et al, 2020) and the results can be used to support further learning.…”

Subsentence Extraction from Text Using Coverage-Based Deep Learning Language Models

“…The importance of personalization in HRI is widely reported and impacts cooperation [2], overall acceptability [3,4] and interaction and engagement [2,5]. Indeed, an early review by Fong [6] on characteristics of successful socially interactive robots suggested that emotions, dialogue, and personality are crucial to facilitating believable HRI.…”