Help test the RF scanner BETA

The map now has a Flock / ALPR camera layer — automated licence-plate readers pulled from OpenStreetMap (the DeFlock project). Turn it on with Layers → “Flock / ALPR cameras” and zoom in. Each camera is a red cone that points the way it faces, so you can see what it's watching.

Keeping it honest — you have to be there. Cities add and remove these cameras, so every one has “✓ Still here” and “✗ Removed” buttons. Tapping them uses your phone's location, and the report is refused unless you're standing at the camera. That's on purpose: nobody can sit at home and mass-report cameras gone (or invent ones that aren't there).

What happens to a report. It's a vote, not an edit — a camera only comes off the map after it's reviewed. And this is where the RF beta closes the loop: when a scanner hears a real Flock camera's wifi next to a mapped one, that camera is marked ✓ confirmed present. So the layer trends toward the cameras that are actually still out there.

Open Termux and paste:

pkg install -y python termux-api
curl -O https://map.sparrowmap.com/rf/phone_scan.py

Then install the Termux:API app from F-Droid if you haven't (it's what lets the script read wifi).

python phone_scan.py            # one scan, right where you are
python phone_scan.py --watch   # keep scanning while you walk or drive

Best test: drive or walk past a spot you know has a Flock or ALPR camera (intersections, town entrances, parking lots). If it broadcasts wifi, you'll see a 🚨 SURVEILLANCE line.

When you're done, it prints a block starting --- copy everything below ---. Copy that and send it to us, along with roughly where you were (a cross street is plenty).

The scanner also flags police in-car equipment — Axon body/dash cameras and docks especially, which are police-dominant. It reports those with a confidence tag alongside the camera finds.

On its own that isn't proof — some in-car gear also lives in work trucks. But mounted in a moving vehicle (a Pi on the dash), hearing patrol gear right where a camera already saw a police vehicle is a strong second signal. That's the goal: RF confirms a visual police sighting, it never names a cop car by itself.

So a Pi in a car does double duty: it maps surveillance cameras as you drive, and it corroborates police sightings along the same route. All of it human-reviewed before anything publishes.

Since 2023 most drones must broadcast Remote ID over the same 2.4/5 GHz wifi and Bluetooth this scanner reads — so it flags drones too (DJI, Skydio, Autel and friends). A drone alone isn't proof of a police drone, and agencies fly the same models everyone else does, so it's treated like the in-car gear: it corroborates when it shows up over a police presence.

A plain scan catches the drone's control link and maker. The full Remote ID broadcast (its id and the operator's location) needs monitor mode — an Alfa adapter on a Pi, or an ESP32 in promiscuous mode — which is the next step. For manned aircraft (police helicopters and surveillance planes) the tool is different: a ~$25 RTL-SDR reading ADS-B on 1090 MHz. That's a separate build; get in touch on Contribute if you want to help with it.

If you have a Pi (or any Linux box) it can scan and upload on its own, no copy-paste. It flags surveillance cameras and police in-car gear; every private network is dropped on the box. Uploads land in a review pen — nothing goes on the public map without a person confirming it.

1. Get the scanner:

sudo apt install -y network-manager python3
curl -O https://map.sparrowmap.com/rf/rf_scan.py

2. Enroll once to get a node id + token (save the token, shown once):

curl -X POST https://map.sparrowmap.com/api/enroll \
  -H "Content-Type: application/json" \
  -d '{"name":"My RF Pi","lat":XX.XXXX,"lon":-YY.YYYY,"kind":"rf"}'

3. Run it (loops, scans, uploads surveillance + police gear only):

python3 rf_scan.py --post --node n_YOURID --token YOURTOKEN

Mount it in a vehicle (best for confirming patrol cars) or a window overlooking a road. A GPS dongle is optional now (position is added in the next version); for a fixed spot the enroll location is enough. A monitor-mode + Bluetooth upgrade (an Alfa adapter) comes later — plain wifi scanning already catches broadcasting cameras and Axon gear.

An ESP32 is a tiny cheap chip that scans wifi and Bluetooth — a matchbox you leave in a car or a window. There's a ready-to-flash sketch: it keeps only surveillance cameras and police gear and uploads them, dropping everything private on-chip, same rule as the rest.

1. Enroll once (from any computer) for a node id + token:

curl -X POST https://map.sparrowmap.com/api/enroll \
  -H "Content-Type: application/json" \
  -d '{"name":"My ESP32","lat":XX.XXXX,"lon":-YY.YYYY,"kind":"rf"}'

2. Flash it with the Arduino IDE (install the "esp32" boards package):

# get the sketch, open it in Arduino IDE
curl -O https://map.sparrowmap.com/rf/esp32_sparrow.ino

Fill in the CONFIG block at the top (the wifi it uploads through, plus your node id + token), then upload. Open the Serial Monitor at 115200 to watch it scan. For a moving sensor, tether it to a phone hotspot. Honest note: the sketch is new and hasn't been proven on real hardware in the field yet — if you flash one, tell us what it catches.

Any device that can see wifi/Bluetooth can feed SparrowMap — another microcontroller, an SDR, a router, whatever you've got. The upload endpoint is open and documented on the Contribute page. Build with it and get in touch; if it works, we'll add a guide for it. The goal is to use any hardware that works, not lock people to one board.