Robotics Research Lab
USC Computer Science
USC Engineering
/ Research / Projects / Data-Driven Interaction Methods for Socially Assistive Robotics: Validation With Children With Autism Spectrum Disorders

Overview Methods People
Publications Support Contact Details

While robots have been used for social interaction, there is great untapped potential for their use as therapeutic social partners. This research provides a process by which a socially assistive robot can be developed and used for socialization therapy for children with autism spectrum disorder (ASD). Toward that end, the robot is designed so that its behavior encourages, facilitates, and trains, social behavior in children with ASD through embodied social interaction. The contributions of this work include a socially assistive robot control architecture, methods for multimodal assistive human-robot interaction (HRI) sensing and recognition, and domain-relevant validation metrics based on existing psychological benchmarks as the basis for the behavior of the robot.

It is generally understood that children with ASD typically respond better, socially and intellectually, to computers and robots than to humans in similar contexts. It has also been observed that robots can inspire social behavior in children with ASD. This work describes a methodology for designing socially assistive robot systems that encourage, through social interaction, a measurable increase in social behavior.


Enabling a robot to understand social behavior, and do so while interacting with the child, is a challenging problem. Children are highly individual and thus technology used for social interaction requires recognition of a wide-range of social behavior. This argues for data-driven methods that capture the relevant range of interactions. This work addresses the challenge of designing data-driven behaviors for socially assistive robots in order to enable them to recognize and appropriately respond to a child's free-form behavior in unstructured play contexts. The focus on free-form behavior is inspired by and grounded in the DIR/Floortime approach to therapeutic intervention with children with autism spectrum disorders (ASD). This approach emphasizes fostering engagement through play, recognizing social behavior and using "engagements" to bolster social interactions.

This research presents a data-driven methodology and a validated experimental framework for enabling fully autonomous robots to interact with both typically developing children and children with ASD in undirected scenarios using socially appropriate behavior, especially where spatial interaction is concerned. Autonomous robot operation as a critical aspect of the methodology; save for safety interventions by a human operator, the robot acts of its own accord. The robot and child engage in free-form interaction, in part though distance-oriented behaviors; the robot must be able to recognize the child's behaviors and respond to them appropriately. This research presents the following computational contributions with therapeutic potential:

The three main contributions above: averse behavior detection, model-based trajectory planning, and data-driven feedback, have been instantiated and validated in several SAR systems using autonomous person sensing, behavior interpretation, and action selection, for the purposes of detecting, provoking, and encouraging both human-human and human-robot social interaction. The validated systems were tested in experiments that evaluated the system design, the accuracy of the robot's ability to interpret observed behavior, the appropriateness of the robot's responses, and the quality of the child-robot and child-parent social behavior interaction. The evaluation experiments were conducted with both children with ASD and typically developing children. The systems were also used to explore the therapeutic potential of socially assistive robots facilitated by the developed models, architecture, and experiment framework.



Journal Articles

Conference Publications


This project was funded in part by the National Science Foundation (IIS-0803565), the Okawa Foundation, the Institute for Creative Technologies, the Dan Marino Foundation through the Marino Autism Research Institute (MARI), the USC Provost's Center for Interdisciplinary Research, and AnthroTronix, Inc.


David Feil-Seifer