Signal Heatmap Layer
Traffic Quadrant Layer
Location Sharing Opt-In
Geo Tagged Images
Geo Tagged Images

CrowdSenseAI

CrowdSenseAI helps university students find the perfect place to study by combining AI-driven crowd prediction with real-time WiFi mapping across campus.

The Problem

Students waste time wandering campus looking for quiet, well-connected spaces to work. There's no reliable way to know where crowds will be or where the WiFi actually holds up.

What It Does

CrowdSenseAI runs two core features on an interactive campus map. The traffic prediction layer uses a neural network trained on anonymized, crowdsourced GPS data to forecast crowd density up to 3 hours in advance. The WiFi layer continuously scans nearby networks as users move through campus, feeding signal strength observations into a Gaussian Process Regression model that predicts connectivity across the entire campus grid. Together, the two layers let students identify spots that are both uncrowded and well-connected. Additionally, the app supports geo-location based file/image upload that provides crucial information about the university campus. (Example: upload wheelchair ramp-doors images and pin a location, that allows visitors with disabilities to find closest ramp doors)

Privacy

User location data is clustered before it's used. The traffic layer displays activity by quadrant, not by individual. Users are prompted to opt in when they first open the app and can opt out at any time. All map data is viewable without sharing your location at all.

Tech Stack

React and Tailwind CSS frontend, Electron for the desktop client, Supabase backend, deck.gl for map rendering, node-wifi for network metrics.

Looking Ahead

The same data infrastructure could power a late-night safety layer, an optimal fundraiser location tool, and an administrative dashboard for campus operations.