Inspiration
Traditional mathematics learning can be challenging for students with dyslexia and learners who struggle with conventional teaching methods. Dense instructions, difficult-to-read text, information overload, and the fear of making mistakes can make learning mathematics frustrating.
We wanted to create a learning environment that adapts to different learning needs rather than expecting every student to learn in the same way.
This inspired us to build Perceptron, an accessibility-focused, gamified mathematics learning application. It combines interactive practice, audio assistance, progressive hints, and a reward-based learning experience to help students understand mathematical concepts at their own pace.
Our primary goal is to make mathematics more approachable, reduce learning frustration, and help students develop confidence through positive learning experiences.
What it does
Perceptron transforms mathematics practice into an interactive and supportive experience, with its initial proof of concept focusing on fractions.
Its key features include:
- Interactive Mathematics Practice: Students solve fraction-based multiple-choice questions through a simple, distraction-free interface.
- Read-Aloud Assistance: Students can listen to questions instead of relying entirely on written instructions.
- Word-by-Word Highlighting: Text is highlighted as it is spoken, helping students follow the content more easily.
- Progressive Hints: Step-by-step guidance helps students understand difficult questions without immediately revealing the answer.
- Supportive Feedback: Incorrect answers encourage another attempt without removing previously earned rewards.
- Honey Reward System: Students earn honey drops for correct answers, encouraging consistent practice.
- Accessible Interface: High-contrast colours, readable typography, clear instructions, and predictable interactions make the experience easier to navigate.
- Planned Dyslexia-Friendly Controls: Customisable font size, letter spacing, word spacing, and motion settings.
- Planned Calming Break Experience: A dedicated space where students can pause, relax, and return to learning.
Perceptron focuses on supporting different learning preferences rather than treating incorrect answers as failures.
How we built it
We developed Perceptron using a frontend-backend architecture that separates user interactions from application logic.
| Technology | Purpose |
|---|---|
| Vue 3 | Builds the interactive frontend and manages application components. |
| Vite | Provides a fast development and build environment. |
| Python | Handles backend application logic. |
| Flask | Manages question delivery, answer validation, hints, and reward calculations. |
| JavaScript Fetch API | Communicates with the backend through HTTP requests and JSON responses. |
| HTML5 | Provides the semantic structure of the application. |
| CSS3 | Implements the colourful neubrutalist interface, visual feedback, and responsive layouts. |
| Web Speech API | Provides browser-based text-to-speech functionality. |
| SQLite | Planned database for persistent learner progress and preferences. |
Development Approach
We developed Perceptron incrementally:
- Established communication between the Vue frontend and Flask backend.
- Implemented the question-and-answer workflow.
- Added server-side answer validation and reward calculations.
- Integrated read-aloud functionality and word-highlighting support.
- Developed progressive hints and answer explanations.
- Tested individual features and refined the learning experience.
This architecture keeps important learning logic on the backend while allowing the frontend to focus on accessibility and user experience.
Challenges we ran into
1. Designing a Non-Punitive Learning Experience
One of our biggest challenges was creating a reward system that motivates students without making incorrect answers feel like failure.
We addressed this through progressive hints and supportive feedback. Students can attempt questions again without losing previously earned honey drops.
2. Synchronising Speech and Text Highlighting
The Web Speech API behaves differently across browsers. In particular, word-boundary events are not consistently supported.
To address this, we implemented word-highlighting functionality with a fallback approach that processes speech one word at a time when reliable highlighting events are unavailable.
3. Separating Frontend and Backend Responsibilities
We needed to ensure that correct answers and reward calculations were not exposed directly through frontend logic.
We addressed this by handling answer validation and reward calculations through Flask endpoints, keeping the frontend responsible for displaying questions and collecting student responses.
4. Balancing Accessibility and Visual Design
We wanted the interface to feel engaging without overwhelming students with unnecessary visual elements.
We focused on clear layouts, readable text, predictable interactions, and visual feedback to make the learning experience more approachable.
Accomplishments that we're proud of
We are proud to have transformed an educational idea into a working application with a clear technical foundation.
Our major accomplishments include:
- Developed a functional Vue 3 and Flask integration.
- Implemented a working mathematics question-and-answer system.
- Kept answer validation and reward calculations on the backend.
- Integrated text-to-speech functionality with a word-highlighting strategy and fallback.
- Built a progressive hint system that guides students towards understanding.
- Designed a reward mechanism that encourages learning without penalising mistakes.
- Created an accessibility-focused interface using a colourful neubrutalist design approach.
- Established an incremental development and testing workflow.
Above all, we are proud that accessibility and learner confidence influenced our product decisions from the beginning rather than being treated as additional features.
What we learned
Building Perceptron taught us that educational technology requires more than presenting questions and calculating scores.
We learned:
- Full-stack development: Connecting Vue 3 with Flask using HTTP requests and JSON.
- Backend architecture: Separating application logic, answer validation, and reward calculations.
- Web accessibility: Designing interfaces that consider readability, audio assistance, and different learning preferences.
- Speech technologies: Working with browser-based speech synthesis and handling inconsistent browser behaviour.
- Gamification: Using rewards and progressive guidance to encourage learning without creating unnecessary pressure.
- Incremental development: Building, testing, and improving individual features instead of implementing everything simultaneously.
Most importantly, we learned that accessibility should be considered throughout the development process, not added as an afterthought.
What's next for Perceptron
We plan to expand Perceptron into a more complete and personalised mathematics learning platform.
Our roadmap includes:
| Planned Feature | Description |
|---|---|
| Persistent Learning Progress | Use SQLite to save learning history, honey rewards, levels, and student preferences. |
| Placement Assessment | Introduce an initial assessment to identify an appropriate starting difficulty level. |
| Expanded Question Bank | Add structured questions, hints, and explanations across three difficulty levels. |
| Dyslexia-Friendly Display | Introduce adjustable font styles, font sizes, letter spacing, word spacing, and line spacing. |
| Reading Assistance | Improve word highlighting, speech controls, and reading customisation. |
| Calming Break Experience | Introduce a relaxing activity that allows students to take breaks and return without losing progress. |
| Accessibility Testing | Evaluate keyboard navigation, screen-reader compatibility, contrast, responsive layouts, and speech behaviour. |
| Mathematics Expansion | Extend learning beyond fractions to percentages, algebra, and other mathematical concepts. |
Our Vision
We envision Perceptron as an accessible mathematics learning platform where students can learn without unnecessary pressure, explore concepts through interactive guidance, and progress at their own pace.
We want to build an environment where making mistakes is part of learning, different learning needs are respected, and every student has an opportunity to develop confidence in mathematics.
Live Demo: Open Perceptron
Built With
- flask
- vue
