Inspiration

Our team's motivation for improving technology comes from wanting to make a difference in people's lives. As college students, we've experienced how ongoing construction on our campuses can make getting to class difficult. This made us think about how people with mobility limitations, such as wheelchair users or people using crutches, deal with blocked sidewalks and inaccessible pathways every day. We wanted to create something that could make navigating these spaces easier for everyone.

What it does

Beacon is an accessibility navigation platform designed to help wheelchair users and individuals with mobility limitations navigate campuses and cities more easily. Unlike traditional navigation apps that prioritize the shortest route, Beacon finds the quickest accessible route based on users' mobility needs, available infrastructure, and community-reported obstacles.

The application combines smart routing, real-time accessibility reports, community contributions, and an AI messaging companion powered by Photon 2.0. Beacon also features a live 3D map and a neon interface designed to be both modern and easy to use.

How we built it

We built Beacon’s web app with Next.js, TypeScript and Tailwind CSS, and deployed it on Vercel, which built a new preview from GitHub on every push. The map uses React Leaflet with OpenStreetMap data. OpenStreetMap was a natural fit because it includes accessibility information such as curb ramps, wheelchair access and surface types. Supabase stores user profiles and mobility preferences along with community-reported obstacles like blocked sidewalks, broken elevators and construction zones, so a new report shows up on the map for everyone right away.

The AI companion runs on Photon 2.0’s Spectrum SDK, which puts Beacon into iMessage. Users can text Beacon to ask for an accessible route, report an obstacle, or check whether a path is clear, without opening an app. The OpenAI API understands each message and writes the replies, and conversation history saved in Supabase lets the companion remember each user’s needs over time. We worked together through Git and GitHub, and used Claude to learn the brand-new Photon SDK quickly and set up our starter agent.

Challenges we ran into

Defining “accessible.” The shortest route and the best accessible route are often very different. We had to decide how to weigh stairs, curbs, steep slopes and reported obstacles, and how to change that for different mobility needs. Gaps in map data. OpenStreetMap’s accessibility information is incomplete, especially for construction that changes week to week. That’s why community reports became a core feature rather than an extra. Learning Photon 2.0 during the hackathon. The SDK launched only weeks ago, so there were very few examples. We had to understand how messages, conversations and channels fit together before we could build on them. Leaflet with Next.js. Leaflet needs the browser, so we had to load the map only on the client side to avoid server-rendering errors. Making the AI sound human. Getting the companion to reply like a helpful friend texting, short and clear, instead of writing paragraphs took a lot of prompt work.

Accomplishments that we're proud of

We have a working iMessage companion: you can text Beacon and get accessibility help in a conversation, right where you already talk to people. Routing reflects each user’s mobility needs, not just distance. Community reports update the map live, so one person’s report helps everyone after them. We built and deployed a full-stack app in one weekend while learning a brand-new SDK. Most of all, we built something we’d want to exist on our own campus.

What we learned

We learned that accessibility goes much deeper than adding a “wheelchair mode.” A route that works for someone on crutches might not work for a power-wheelchair user, so good design starts with listening to real needs. On the technical side, we learned to build a messaging agent with Photon, to work with geographic data and real-time updates, and to combine an AI model with stored memory so the agent can give personal, context-aware help. We also learned to divide work and merge quickly as a team under time pressure.

What's next for Untitled

We hope to integrate Beacon into the UT Dallas campus first, helping students and visitors navigate construction zones and find accessible pathways. From there, we want to expand to other universities and cities that prioritize walkability and inclusivity. Eventually, we hope Beacon can be used across cities worldwide, making navigation easier and more accessible for everyone.

Built With

Share this project:

Updates

Submission history