Play.Care by Harimata

The Challenge: Early Autism Detection

Autism Spectrum Disorder (ASD) is present from infancy, yet accurate clinical identification currently requires extensive expert diagnostic training. We addressed this significant barrier by focusing on a groundbreaking computational marker: new evidence suggesting that disruptions in motor timing and integration may underpin the disorder.

Groundbreaking Research & Computational Assessment

Our team conducted a large-scale series of studies, analyzing the motor patterns and gesture forces of over 600 children aged 3–6, including those with autism, other developmental disorders, and typically developing children.

We utilized standard smart tablets—leveraging their touch-sensitive screens and embedded inertial sensors—to capture granular, quantitative movement data during structured gameplay.

Key Findings:

• Machine learning analysis of children’s motor patterns identified autism with an accuracy of up to 93%.
• The analysis revealed a distinct autism motor signature including greater forces at contact, faster/larger gesture kinematics, and more distal use of space.

This work validates movement disruption as a fundamental aspect of ASD, pioneering a new path for objective computational assessment via accessible, engaging smart device gameplay.

The Solution: Play.Care App

Play.Care is the pioneering medical tablet application designed for the early, objective detection of autism in children aged 3–6. The app integrates research and clinical utility by:

1. Engaging Assessment: Children complete a brief, 7-minute test embedded within specially designed educational games.
2. Advanced Analysis: Machine learning algorithms scrutinize the child’s fine motor patterns and movement kinematics in real-time.
3. Objective Reporting: The child's data is compared against the established autistic motor signature, generating a detailed report outlining the risk of autism for clinical follow-up.

My Role (Research Lead)

As the Research Lead, I drove the project from concept to rigorous clinical validation, engaging in every critical facet of the app's development and implementation:

• Securing project funding and managing research finances.
• Orchestrating the entire research process, including study design and data analysis.
• Crafting and designing the educational games used for the 7-minute assessment.
• Designing and testing user experiences (UX/UI) for the app.
• Planning and overseeing Phase 3 clinical studies to rigorously validate the app's efficacy. 

Engaging gameplay

During a brief 7-minute assessment, children engage with specially designed educational games on a smart tablet.

Advanced analysis

The app captures and analyzes their movement patterns using advanced machine learning algorithms. By comparing these patterns with those of children diagnosed with autism, Play.Care can identify potential signs of the disorder with an accuracy of up to 93%

Why it matters?

Our research has shown that disruptions in motor timing and integration are key indicators of autism. By focusing on these movement patterns, Play.Care not only offers a new avenue for early identification but also empowers parents and caregivers with a detailed report on their child's developmental risk.