AccessAudit

AI-powered wheelchair accessibility auditor using NVIDIA Nemotron vision models.

How It Works

AccessAudit runs a 4-step autonomous agent loop for any address:

1. Geocode — Converts the address to coordinates via OpenStreetMap Nominatim
2. Capture — Fetches 4 Google Street View images facing N, E, S, W
3. Analyze — Sends each image to nvidia/nemotron-nano-12b-v2-vl for accessibility analysis (sidewalks, curb cuts, ramps, obstacles, surface quality)
4. Synthesize — nvidia/nemotron-3-super-120b-a12b combines all findings into a scored report with an A–F grade and actionable recommendations

Tech Stack

• Framework — Next.js 14 (App Router, TypeScript)
• Styling — Tailwind CSS
• AI Vision — NVIDIA Nemotron Nano 12B v2 VL
• AI Synthesis — NVIDIA Nemotron Super 120B
• Maps — React-Leaflet + OpenStreetMap tiles
• Geocoding — Nominatim (free, no key required)
• Street View — Google Street View Static API

Setup

1. Clone the repo and install dependencies:

   npm install

2. Create .env.local in the project root:

   NVIDIA_API_KEY=nvapi-...
   GOOGLE_MAPS_API_KEY=AIzaSy...
   

3. Run the dev server:

   npm run dev

4. Open http://localhost:3000 and enter any address.

Social Impact

Over 1.3 billion people worldwide live with some form of disability — roughly 1 in 6 of the global population. For wheelchair users, inaccessible sidewalks, missing curb cuts, and broken ramps aren't inconveniences; they are barriers to employment, healthcare, education, and daily life.

AccessAudit makes accessibility auditing instant and scalable. What once required a trained inspector and hours of fieldwork can now be done in under a minute for any address on Earth — helping city planners, disability advocates, and businesses identify and prioritize accessibility improvements before people encounter them in the real world.