A thumbnail for the project Empowering Speech showing the name of the project and a 3D model of the proposed idea showing a wheelchair with a small screen attached to the right handle, and a person sitting on the wheelchair with an EEG headset connected to the screen.
TYPE OF WORK
Team project
CATEGORY
Disability Interaction / Speculative Design
TIME
Feb 2024 - Mar 2024

Overview

The core of this project's module was shaped by four intensive, three-hour co-design sessions with individuals with various disabilities. In these sessions, we worked closely with each participant to identify specific challenges, brainstorm solutions, and refine one concept into a viable assistive technology design.

This particular project highlights one of those sessions with an individual with aphasia, where we aimed to create an innovative technology that supports relationship maintenance by facilitating self-initiated communication. Traditional speech-generating devices often feel impersonal, so this design prioritizes retaining the user’s natural voice and communication style.

My role in this project involved being the facilitator during the co-design sessions to enable a space for inclusive and creative participation with our guests, as well as leading ideation during the sessions.

Problem Statement

Aphasia, often a result of brain injury or stroke, impacts the ability to communicate, creating barriers that can strain relationships. Conventional assistive devices for aphasia may help with basic communication but lack the personal touch that close relationships require. This project sought to address the unique challenge of helping individuals initiate speech while preserving the authenticity of their voice, focusing on how technology can aid self-cueing for speech initiation without erasing personal identity.

Co-design & Requirements

During the co-design session with the participant, their partner, and researchers, we explored the participant’s communication needs. While able to speak full sentences, the participant struggled with initiating speech, a frustration that impacted their relationship. Insights from the participant’s speech therapist revealed that starting with a phonemic or orthographic cue greatly improved the ability to continue speaking.

The participant and their partner emphasized the importance of preserving the participant’s natural voice over using synthetic speech-generating devices, as this maintained identity and connection.

This led us to identify three key requirements:
  • Enable self-cueing to support independence.
  • Use the participant’s own words rather than predictive text.
  • Preserve the participant's natural voice.

Conceptual Design

To address these needs, we explored the use of EEG headsets as a potential solution. EEG headsets can capture brain signals, translating the user’s intent into actions on a connected device. By detecting the user’s intention to speak, the headset could trigger a cue on an external display, helping the participant initiate speech independently. This speculative concept aimed to meet the participant’s core requirements by supporting authentic, self-initiated communication without replacing their own voice.
A 3D model showing the conceptual design. Image shows an individual on a wheelchair. The individual has an EEG headset on their head and the wheelchair has a medium sized display on the right handle such that the individual can interact with it.

Evaluation Plan

Due to technical limitations, a full evaluation wasn’t feasible. However, a two-stage plan was designed for future testing of the system’s usability and impact.
Stage 1: Controlled Setting Evaluation
In a lab setting, participants would attempt to speak specific phrases with system assistance. This initial test would gauge usability, error rate, and comfort with the EEG headset, identifying any early obstacles in using the system effectively.
Stage 2: In-the-Wild Study
In the second stage, participants would use the system at home over several weeks to observe real-life impact on communication and relationship dynamics. Follow-up interviews would provide insights into how well the system supports authentic, self-initiated communication.

Limitations

While this design is promising, it has several limitations. First, it was developed based on the input from a single participant, meaning its effectiveness and relevance may vary widely across different individuals with aphasia. For broader applicability, further testing with a diverse participant pool is necessary.

The use of an EEG headset could raise privacy concerns, as brain signals are highly sensitive and require rigorous regulatory measures, especially for a vulnerable user group. Additionally, the device's visible design could be stigmatizing, as many users prefer assistive technologies that are discreet in public settings.

Finally, the cost of such product could limiting accessibility for many users. Financial constraints, including lower earnings among people with disabilities, make affordability an essential consideration for real-world adoption. Nonetheless, this project encourages ongoing innovation to bridge low-tech and high-tech solutions, making assistive technology more inclusive.

Project Relevance

This project highlights how assistive technology can go beyond functional support to foster connection, self-expression, and emotional well-being.

It addresses a key gap in assistive technology for individuals with aphasia, focusing on enhancing their ability to initiate speech with close ones, an often overlooked aspect of disabilities.

Its relevance extends to the broader field of inclusive design, illustrating how advanced technology can be thoughtfully adapted to meet real human needs, promoting inclusivity in the design of future assistive devices.

More Projects

A thumbnail for the project SustainaBear showing the name of the project and three mockups of the productA thumbnail for the project MediFlora showing the name of the project and three mockups of the productA thumbnail for the Masters Dissertation showing a snippet of the academic paper
VIEW ALL WORK