A crowdsourced platform for mapping the Sub-GHz radios in the world around you — a “Wigle for RF”. Upload a .sub/.fff capture with GPS, let it auto-identify the device against known signature databases, and watch it land on an interactive map. Built in the open by OptinAmpOut.
Every garage remote, weather station, TPMS sensor and smart doorbell is shouting on the 300–900 MHz bands. GigLez turns those captures into a map: submit a signal file plus a GPS fix from any capture tool, and the platform parses it, identifies the device, and plots it — building a crowdsourced atlas of the invisible radios around us.
A submission carries a signal file, GPS coordinates and a timestamp. From there it’s automatic.
A .sub/.fff file with required GPS (lat/lon) and timestamp — single, or a ZIP batch with a manifest.json.
Protocol, frequency, modulation and bit patterns are pulled straight from the raw capture.
Compared against Flipper Zero / RTL_433 signature databases with exact, partial and pattern-based confidence scoring.
The device lands on an interactive map and heatmap; the community can confirm or correct the auto-ID.
GPS markers for every capture plus a density heatmap by geographic area — pan, zoom and explore coverage.
By device type, frequency, protocol, date range or geographic bounding box — with export.
Total captures, unique devices and geographic coverage — the health of the dataset at a glance.
Matches are ranked by similarity — exact, partial and pattern-based — not a blind yes/no.
Users confirm or correct automatic identifications — the dataset gets sharper over time.
Top contributors by uploads and verifications — the crowdsourcing flywheel that made Wigle work.
GigLez is device-independent — if your tool outputs .sub or .fff, it works.
The native .sub capture format — the most common Sub-GHz field tool.
Cheap software-defined radio dongles — capture and submit from a laptop.
Wide-band SDR for the serious operator — full transmit/receive coverage.
Compact ESP32-based RF boards — pocketable, GPS-friendly capture.
Built to dedupe by file hash and GPS proximity, store both raw files and parsed metadata, and index by frequency, protocol, location and time.
| Layer | Stack |
|---|---|
| API | FastAPI — async submission & query endpoints. |
| Datastore | PostgreSQL + PostGIS via SQLAlchemy / GeoAlchemy2, migrations with Alembic. |
| Capture I/O | pyserial / aioserial for live device reads. |
| Signal decoding | An RF spectrum analyzer & decoder in JavaScript, plus a Python parsing engine. |
| Signatures | Flipper Zero & RTL_433 device databases for auto-identification. |
| Mobile | python-for-android path for on-the-go capture & submission. |
GigLez ships with a full test suite, geospatial migrations and a safe-install path that creates a virtualenv, checks for conflicts and can roll back. It’s the kind of data-heavy platform work we take on for clients — built in the open here.
GigLez is geospatial ingestion, signal parsing and crowdsourced mapping in one system — the kind of full-stack data product we build for clients. Tell us what you’re mapping.
Book a build call →