Robot Mediated Autism Intervention: Hardware, Software, and Curriculum

Organizers: Momotaz Begum and Michael Radice

More than a decade of research in human-robot interaction (HRI) has established the fact that embodied robots have a huge potential as an intervention tool for some children with autism [1,2,3] . Currently 1 in 68 children in the USA are affected by autism [4]. The average cost for a family to care for an autistic child is approximately $60K/year. World-wide prevalence of autism triggered the need for training professionals and developing new technologies to efficiently deliver autism intervention. Technology-aided intervention are considered as an emerging evidence-based practice (EBP) in autism and physically embodied robots are being cited as an example of emerging tools for autism intervention [5]. In the recent years clinical use of commercially available robots (e.g. NAO from Softbank Robotics, Milo from Robokind) in autism intervention have increased significantly. For example, it is estimated that over 400 robots are currently being deployed in school/therapeutic settings for providing intervention to autistic children [unpublished data, collected based on personal communication with ROBOTTECA.com and Robokind]. Despite these impressive new developments in robot-mediated behavior intervention (RMBI), there is a huge gap in understanding among robotics researchers, robotics industries and stakeholders (autistic children, parents/caregiver, and clinicians) on the clinical utility of robots in autism intervention, best practice in RMBI, and the robotics technology (hardware and software) required to establish RMBI as an effective evidence-based practice (EBP) in autism. This workshop is an effort to bridge that gap. The workshop aims to bring people from academia (both from robotics and clinical sciences) and robotics industry to foster a discussion on how to collaborate on effectively developing robot hardware, novel algorithms to design intelligent robot behaviors required for RMBI, developing curriculum and progress measurement tool for the practitioners to meet the ultimate goal of advancing the social and learning skills of autistic children, helping them secure a confident and comfortable position in life as a child and an adult. The workshop, therefore, will focus on the following two themes:

Best practice in RMBI: Creating a common cause for identifying ever better fact-based tools, strategies, and methodologies to establish robot-mediated intervention as an effective EBP in autism. The workshop will discuss on the possibility of establishing a framework for developing a certification standard for RMBI. Such an effort needs to be backed by clinical researchers on autism, practitioners who deliver RMBI and robotics companies that provides curriculum along with the robot. Some specific points of interest include

What are the existing and potential models of robotic platform functionality that currently or would potentially offer the most promising result in autism intervention?
What type(s), styles of robot behaviors, interactions, games, reward sessions would provide the widest range and most promising engagements for creating and advancing positive results in learning new skills?
What augmented session management technologies for recording, progress management and oversight would help adoption and use of robots in clinical settings?

Hardware and Software for RMBI: Discussing research challenges in the design and development of robot hardware and software (algorithm, language, interface) that are needed to be addressed to enable seamless integration of robots in existing clinical practice in autism. The workshop will discuss the expectations of practitioners and robotics companies from academic researchers in robotics on fundamental development in robot hardware and software that could significantly lower the barriers for robots to enter in clinical practice. The workshop will also discuss about the possibility of initiating industry-led open challenges to solve fundamental problems in robot design (hardware/software) for academic researchers.The specifics of such challenges could be defined by clinical researchers and autism practitioners who encounter issues while using the existing robotic technologies in autism intervention.