AusCare

An Assistive AR Device for Children on the Autism Spectrum

A Project from the University of Modena and Reggio Emilia (UNIMORE)
Dipartimento di Ingegneria "Enzo Ferrari"

Our Mission

Making Therapy Accessible

Autism Spectrum Disorder presents unique challenges, especially in social interaction and emotion recognition. While therapy is crucial, it can be prohibitively expensive, often costing €50-100 per session.

Our goal with AusCare is to leverage technology to create a more affordable, accessible, and effective therapeutic tool. Our prototype was developed for under €250, a single cost that can be driven even lower with production at scale.

AusCare device diagram
-

Core Features

👓 The AusCare Device

The core of AusCare is a custom-built, Raspberry Pi-powered headset. It's designed to be a non-intrusive tool for children, equipped with a camera, a display, and wireless connectivity to support our software features.

😂 Real-Time Emotion Recognition

The device captures the user's view, analyzes facial expressions of people they interact with, and provides a simple emoji on the OLED display to help the child identify the emotion in real-time. This is designed to improve social understanding and interaction skills.

📊 Digital Twin & Data Analytics

Every interaction is logged, creating a "digital twin" of the child's progress. This data is presented in an intuitive web dashboard for parents and therapists to track progress, identify patterns, and personalize treatment plans effectively.

System Architecture

Proof of Concept (PoC)

Our initial prototype integrates most components directly onto the Raspberry Pi. A Flask web server runs locally, and we use NGROK to create a secure tunnel for remote access by doctors or parents, allowing them to view the data dashboard from anywhere.

A Closer Look at the Features

Emotion Recognition in Action

The device uses a camera and a cloud-based API to analyze faces. The recognized emotion is then mapped to a simple emoji and displayed on the headset's OLED screen, offering immediate visual feedback to the child.

Emotion Recognition

The Digital Twin Dashboard

All collected data feeds into a secure, user-friendly dashboard. Here, therapists and parents can view activity logs, track emotional responses over time, and see how many people the child interacted with, providing invaluable quantitative feedback on the therapy's effectiveness.

AusCare Metrics Dashboard

A Scalable Cloud Solution

Scalable Cloud Architecture

A More Scalable Future

For a production-ready system, we envision a decoupled architecture. Key components like the emotion detector, database, and web interface would be hosted on the cloud. This ensures the system is robust, secure, and can handle data from many users simultaneously, while the local device remains lightweight and efficient.

The Future of AusCare

Expanding the Horizon

The AusCare platform is built for growth. Future versions will incorporate features like attention tracking for optimizing training sessions, AR-based educational games, and even non-invasive EEG integration to get a deeper understanding of a child's cognitive state for truly personalized learning tracks.

Our vision is to cover more aspects of the autism spectrum, providing tools not just for emotion recognition, but also for sensory issues, fixations, and learning impairments.

Demo