PictoMaker is an application designed to support Augmentative and Alternative Communication (AAC) by combining pictograms with modern artificial intelligence techniques. Its goal is to make pictogram-based communication faster, more intuitive, and easier to integrate into real-world contexts such as education, therapy, and everyday communication.
This project originated from both a technical and human need: while pictograms are a powerful tool for cognitive accessibility, working with them is often manual, slow, and difficult to scale. PictoMaker addresses this gap by introducing prediction, automation, and language-aware assistance into the pictogram workflow.
Context and Motivation
AAC systems play a crucial role for people who cannot rely on spoken language, including individuals with intellectual disabilities, autism spectrum conditions, cerebral palsy, or other communication challenges. Among these systems, pictograms are widely used due to their clarity and universality.
However, existing tools typically require users or caregivers to manually search, select, and arrange pictograms, which can be cognitively demanding and time-consuming. This limits autonomy and slows down communication.
PictoMaker was created to explore how AI and NLP can act as assistive layers, reducing friction while preserving user control and interpretability.
What PictoMaker Is
PictoMaker is a predictive pictogram keyboard and communicator that assists users while they build messages. Instead of starting from scratch each time, the system provides intelligent suggestions that adapt to context and user intent.
At a high level, PictoMaker:
- Assists users in selecting pictograms more efficiently
- Uses language-aware models to predict likely next elements
- Supports clearer, more fluid message construction
- Prioritizes accessibility and cognitive simplicity over raw automation
The system is designed as a supportive tool, not a replacement for human communication or decision-making.
Technical Approach
PictoMaker combines several technical components:
- Natural Language Processing (NLP) to analyze linguistic structure and context
- Predictive language models to suggest relevant pictograms dynamically
- A modular architecture that allows iteration and experimentation without locking into a single model
- A Python-based implementation focused on clarity, maintainability, and extensibility
Special care was taken to ensure that predictions remain interpretable and that the system can be adapted to different user profiles and communication needs.
From Application to Community
PictoMaker was presented publicly in the talk "Beyond Words: Breaking Communication Barriers with Python" at PyConES 2025, where it generated strong interest and discussion around the intersection of AI, accessibility, and social impact.
The feedback from the community — including educators, developers, and accessibility advocates — reinforced the relevance of the problem and opened conversations about extending PictoMaker in collaboration with professionals in AAC, special education, and related fields.
Impact and Future Directions
PictoMaker demonstrates how applied AI can be used responsibly to:
- Reduce barriers to communication
- Empower users with assistive, not intrusive, technology
- Bridge engineering, accessibility, and real social needs
Future work focuses on:
- Collaborating with AAC professionals and educators
- Evaluating the system in real educational and therapeutic contexts
- Iterating on prediction strategies to better adapt to individual users
The project remains intentionally open-ended, prioritizing real-world validation and ethical deployment over premature scaling.
Related Work
- Talk: Beyond Words: Breaking Communication Barriers with Python (PyConES 2025)
- Article: Building PictoMaker: accessible communication with Python and AI
PictoMaker is an ongoing exploration of how technology can support communication with care, rigor, and impact.