JMIR Publications

ABSTRACT

Background:

Social engagement is critical for the developmental progress of children with autism spectrum disorder (ASD). While tangible technologies and serious games (SGs) show promise, there is a pressing need for a dual-purpose platform that both facilitates collaborative interaction and enables sensing engagement. Existing tangible interfaces often lack integrated sensing capabilities, while devices designed for complex input are typically not optimized for the unique sensory and therapeutic needs of this population.

Objective:

This study presents InteractiSense, a tangible prototype designed to support collaborative play while enabling engagement estimation for children with ASD through embedded multi-modal sensing. Our primary objectives were twofold: (1) to evaluate the prototype’s preference, usability, and workload, and (2) to assess the feasibility of its embedded sensor data in generating valid inputs for machine learning-based engagement classification.

Methods:

We developed InteractiSense, a spherical, soft-textured prototype embedded with seven physio-behavioral sensors, alongside a collaborative SG (“Treasure Hunters”) promoting key social skills. A within-subjects study with 15 children with ASD collected preference, usability, and workload ratings, supplemented by caregiver feedback. To validate sensing feasibility, we analyzed the signal-to-noise ratio (SNR) of the sensor data and trained a preliminary XGBoost model to classify binary engagement states (engaged vs. not-engaged) using video-coded ground truths.

Results:

InteractiSense was perceived favorably, showing a tendency for higher user preference while demonstrating usability and workload metrics comparable to the familiar baseline (Xbox controller). Caregivers noted the prototype’s distinct physical features—such as its round shape, soft texture, and simplified configuration—as contributing to this favorable perception by children with ASD. Signal quality analysis indicated the embedded sensing modalities provided analyzable data. Finally, a preliminary XGBoost model achieved 83% balanced accuracy in binary engagement classification, exceeding random and majority-class baselines.

Conclusions:

These findings provide initial evidence that InteractiSense can serve as a viable dual-purpose platform. It functions as an accessible and preferred interaction medium for collaborative and social SGs while simultaneously enabling objective engagement and behavioral monitoring by capturing valid, multi-modal data. This work demonstrates the feasibility of embedding such engagement-monitoring capabilities directly into tangible interfaces, paving the way for more adaptive, content-rich therapeutic and educational tools for children with ASD.