Spatiora: Privacy-First Spatial Intelligence Through Wi-Fi
Link to open source: https://github.com/saaheerpurav/spatiora
What is Spatiora?
Spatiora explores how ordinary Wi-Fi signals can help us understand physical spaces without cameras or wearables. Three low-cost ESP32 boards capture changes in Wi-Fi Channel State Information, which our software processes to estimate presence and motion. A live 3D dashboard visualizes the sensing area and signal activity.
Our prototype connects sensor readings to Google Cloud through Cloud Run and Pub/Sub, with BigQuery and Cloud Storage supporting analysis and historical records. AI-assisted signal interpretation helps us explore patterns in the readings.
Why we built it
Cameras can provide valuable information, but they also raise privacy concerns in homes, care facilities, and shared spaces. Wearables require people to remember to wear and charge them. We wanted to explore an affordable alternative that senses changes in an environment without recording identifiable images.
How it can help others
Spatiora could support privacy-conscious occupancy monitoring, activity alerts for caregivers, and better use of shared rooms and buildings. Over time, we aim to extend it toward reliable location estimation and detection of unusual activity.
The current build is an experimental prototype. Presence accuracy depends on placement and calibration, while precise 3D localization, posture recognition, and fall detection remain future development goals.
Explore Spatiora on GitHub:
https://github.com/saaheerpurav/spatiora
This build was uploaded as a hackathon project
