Designing a Pictogram-Based Visual Language Tool for Self-Determination in Individuals with Complex Communication Needs: A Practice-based Research Approach

Abstract

This research focuses on developing PictoNet, a pictogram-based visual language system to support individuals with complex communication needs (CCN) in Chile and Aotearoa New Zealand. The project aims to improve self-determination and accessibility by creating a tool that connects textual and pictographic communication through an AI-assisted system. By embracing a practice-led research approach, the study combines visual communication, interaction design, and artificial intelligence to create an adaptive pictographic framework. It builds upon the PICTOS project in Chile, which addresses cognitive accessibility in public services through pictographic sequences and expands towards communication sovereignty within a universal design framework. The methodology includes collaboration with practitioners and Alternative and Augmentative Communication (AAC) specialists and ethnographic fieldwork, interviews, and usability studies to evaluate existing pictographic systems. In PictoNet, artificial intelligence acts as a design material to help generate pictograms in AAC, allowing real-time adaptation to user needs while ensuring transparency and reversibility. PictoNet is envisioned as an open-source, federated system where collective contributions enrich the pictogram library, promoting collaborative visual communication. In addition to its technical contributions, this study examines the legitimacy of practice-led research within the academic framework, advocating for its significance alongside traditional scientific methods. The research emphasises design as a fundamental aspect of inquiry by recognising the importance of design knowledge and the complexities of visual language systems. Ultimately, this study aims to enhance inclusive communication and offers a new approach to augmentative and alternative communication in a world increasingly influenced by AI.