Berkeley MDes

Gracy Kureel

MDes 2023

Advancing Autism Care Through Wearable Physiological Sensing

Introduction

This study addresses the critical need for innovative communication methods for individuals with profound autism, characterized by severe cognitive impairments, very low IQ, and significant communication challenges. Traditional communication strategies often fall short, increasing caregiver burden and the risk of mental health issues. The research explores non-verbal communication through physiological and emotion-based analyses, focusing on wearable physiological sensing and affect-based detection systems. These tools are hypothesized to effectively convey basic needs, like thirst, in those with profound autism.

Process

The inception of this project was driven by a desire to develop technology that significantly enhances communication for individuals with profound autism. The journey began with consulting professionals in allied fields and conducting a comprehensive review of existing literature, which highlighted critical concerns such as privacy issues, data quality, and the practicality of current methods. These insights led to the formulation of research questions focused on capturing naturalistic non-verbal communication, improving data quality with contextual insights, developing personalized machine learning methods, and crafting an engaging Augmentative and Alternative Communication (AAC) interface.

Simultaneously, I embarked on developing the user interface for the AAC platform. Initial designs using bar graphs were transformed into a more intuitive format through peer discussions, emphasizing the need for clear and actionable information for caregivers. The app’s final design focused on delivering easy-to-understand visualizations and notifications, optimizing for user-friendliness and effective communication.

Final Design

The final design of the project, named ‘Aiden’, centers around
an AAC platform tailored for individuals with profound autism. The development process unfolded in stages, starting with extensive data collection using the Empatica EÉ sensor to capture physiological changes in neurotypical individuals in both dehydrated and hydrated states. This data collection was crucial for establishing the groundwork of the project.

This led to the preliminary development of the AAC platform using Android Studio and the Empatica EÉ’s SDK. The prototype application, designed with caregivers in mind, provides real- time alerts and notifications based on the machine learning model’s analysis. The design was informed by the data collected and analyzed, ensuring it was responsive to the specific needs identified in individuals with profound autism.

The current focus of ‘Aiden’ includes addressing fundamental needs like thirst, hunger, pain, and the urgency to urinate. These needs were selected due to their immediacy and impact on the quality of life of individuals with profound autism. By effectively monitoring and responding to these needs, ‘Aiden’ aims to provide a more comprehensive and caring environment for its users.

Full listing