Voice City

Inspiration

It began with a simple truth — the world isn’t built equally for everyone. We move through our cities with ease, but for millions, every step, every bus ride, and every doorway is a challenge. A step too high. A sign unreadable. A route without guidance. We met people whose strength was hidden behind quiet frustrations — the ones who just wanted to belong in a world that often forgets to include them. That’s when we realised: accessibility is not a privilege — it’s a right. VoiceCity was born from empathy and innovation — a vision to make cities listen to their people. We wanted to give every individual, regardless of ability, a voice that shapes the spaces they move through. Our goal is simple yet powerful: To build cities that speak, adapt, and include. To turn technology into an instrument of equality, powered by community and compassion.

What it does

VoiceCity is an AI-enhanced accessibility mapping application that allows users to review, navigate, and verify accessible public spaces across the UK.

✨ Key capabilities include:

🧭 Community-Rated Accessibility: Users can rate spaces based on mobility, hearing, and vision accessibility.
🗺️ AR & Voice-Guided Navigation: Provides real-time, step-by-step directions that adapt to each user’s needs.
🤖 AI-Powered Summaries: Automatically generates accessibility summaries from community reviews.
🎨 Inclusive UI Design: Features voice commands, high-contrast visuals, and haptic feedback for all users.
🏅 VoiceCity Verified Badge: Awarded to businesses that meet accessibility standards.

How we built it

We developed VoiceCity as a concept prototype using:

🧠 Google AI Studio – for website design and AI-based summarization
💻 Flask & HTML/CSS/JS – to build the web interface
☁️ Render (Free Tier) – for deployment and hosting
🔗 GitHub – for collaboration and version control

The workflow involved:

⚙️ Defining accessibility categories and rating criteria
🎨 Designing the UI to support both visual and non-visual accessibility
🤖 Integrating AI summarization for user-generated content

Challenges we ran into

⏺️ Handling the diversity of accessibility needs (mobility, vision, hearing)
⏺️ Designing an interface that’s inclusive yet simple
⏺️ Integrating AI models efficiently with limited resources
⏺️ Managing deployment limitations under Render’s free hosting tier
⏺️ Ensuring the data model stays ethical, unbiased, and community-centered

Accomplishments that we're proud of

🏆 Built a fully functional prototype website for VoiceCity
🏆 Represented the University of Leeds at AETHRA GLOBAL IDEATHON 2025
🏆 Developed a scalable model that aligns with UK Equality Act 2010 standards
🏆 Designed a verified accessibility badge system to promote inclusivity in businesses
🏆 Showcased the impact of AI + community participation in solving real-world accessibility challenges

What we learned

📚 The importance of empathy in tech design — accessibility is a right, not an afterthought
📚 Hands-on experience integrating AI summarization with human-centered data
📚 Collaboration, communication, and adaptability under tight ideathon deadlines
📚 Understanding how data-driven insights can empower civic and social change

What's next for AETHRA GLOBAL IDEATHON 2025

👉 Integrate Google Maps API and Transport for London (TfL) data for live accessibility updates
👉 Partner with UK councils, NHS digital inclusion programs, and charities like Scope
👉 Introduce gamified contributions (e.g., badges for verifying 50+ locations)
👉 Enhance predictive AI to suggest accessible alternatives nearby
👉 Expand globally to create a universal accessibility ecosystem