Julight

A Monkey C (Garmin Connect IQ) app that turns a fenix 7X Solar into a BLE central which connects to a nearby ESP32 (BLE peripheral), streams sensor data to it over a GATT link, and reads signal strength (RSSI) as a rough proximity/distance measure.

What it does (Option 2 — GATT data link + distance)

The watch:

1. Scans for the ESP32, matched by a custom service UUID.
2. Reads RSSI from the scan result → crude log-distance estimate shown on screen.
3. Connects (pairs) to the ESP32.
4. Enables notifications on the notify characteristic.
5. Writes sensor payloads to the data characteristic (button press sends a test packet today; the 25 Hz accelerometer / heart-rate batching plugs in next — see Roadmap).

Note: a Garmin watch can only act as a BLE central — it cannot advertise as a peripheral. That's why the other end must be an advertiser (ESP32/nRF/beacon), not a second Garmin watch.

BLE contract (must match the ESP32 firmware)

Role	UUID
Service	c3a1f200-9b0e-4f1a-8a01-0e1d2c3b4a59
Data (watch→ESP32, write)	c3a1f201-9b0e-4f1a-8a01-0e1d2c3b4a59
Notify (ESP32→watch)	c3a1f202-9b0e-4f1a-8a01-0e1d2c3b4a59

The ESP32 must advertise the service UUID so the watch can find it, expose the data characteristic as writable, and the notify characteristic as notify (with a 0x2902 CCCD descriptor).

Project layout

manifest.xml          app id, fenix7x product, BLE + Sensor permissions
monkey.jungle         build config
source/
  JulightApp.mc       app entry; creates the BleManager
  BleManager.mc       scan / connect / RSSI / GATT read-write (the core)
  JulightView.mc      shows state, RSSI and estimated distance
  JulightDelegate.mc  SELECT = send test packet, BACK = exit
resources/            strings + launcher icon
build.sh              build / run helper

Building

Requires the Connect IQ SDK and a developer key (kept outside the repo at ~/.garmin_keys/developer_key.der).

./build.sh        # compile to bin/Julight.prg
./build.sh run    # compile + launch the simulator

Or use the Garmin Monkey C VS Code extension (F5 to run).

Roadmap

• BLE scan + connect + RSSI/distance display
• Write test payload to the ESP32 over GATT
• Register the 25 Hz accelerometer listener and batch samples
• Add heart-rate + stream batched sensor data to the ESP32
• RSSI smoothing / calibration for a more stable distance estimate
• Reconnect / power handling
• WiFi audio mode: stream MP3 FFT (bass) → light brightness over UDP (see python/)