ESP32 Guides

Practical guides for connecting ESP32 microcontrollers to RosFit — from simple MQTT sensors to micro-ROS motor controllers and mesh networks.

Smart greenhouse sensor node

Build an environmental monitoring node that reports temperature, humidity, soil moisture, and light levels to RosFit.

What you'll build:

• ESP32 with DHT22 (temp/humidity), capacitive soil moisture sensor, and BH1750 light sensor
• Connect to RosFit via MQTT over Wi-Fi with token authentication
• Publish sensor readings every 30 seconds to separate telemetry topics
• Configure dashboard gauges for each metric
• Set up threshold alerts (temperature > 35°C, soil moisture < 20%)

Hardware: ESP32-WROOM-32, DHT22, soil moisture sensor, BH1750, breadboard.

#include <WiFi.h>
#include <PubSubClient.h>
#include <ArduinoJson.h>

const char* mqtt_server = "emqx.rosfit.local";
const int mqtt_port = 1883;
const char* device_id = "sensor-greenhouse-01";
const char* mqtt_token = "mqtt_tok_greenhouse01";

WiFiClient espClient;
PubSubClient client(espClient);

void publishTelemetry() {
  JsonDocument doc;
  doc["temperature"] = readDHT22Temp();
  doc["humidity"] = readDHT22Humidity();
  doc["soil_moisture"] = readSoilMoisture();
  doc["light_lux"] = readBH1750();
  doc["timestamp"] = millis();

  char buffer[256];
  serializeJson(doc, buffer);

  String topic = String("rosfit/") + device_id + "/telemetry/environment";
  client.publish(topic.c_str(), buffer, true);
}

micro-ROS motor controller

Run micro-ROS on an ESP32 to control DC motors through the ROS 2 graph, bridged to the cloud via RosFit.

What you'll build:

• ESP32 running micro-ROS agent over serial or Wi-Fi
• Subscribe to geometry_msgs/Twist on /cmd_vel for differential drive control
• Publish encoder-based odometry on /odom
• Bridge both topics to RosFit cloud via a co-located micro-ROS agent
• Control the robot from the dashboard joystick widget

Hardware: ESP32, L298N motor driver, DC motors with encoders.

Key topics covered: micro-ROS setup, XRCE-DDS agent, Twist message handling, encoder odometry.

# Flash micro-ROS firmware
ros2 run micro_ros_setup create_firmware_ws.sh freertos esp32

# Start the micro-ROS agent
ros2 run micro_ros_agent micro_ros_agent serial --dev /dev/ttyUSB0

# Register the device
rosfit devices add --name "esp32-motor-01" --type micro_ros_mcu --group lab

Battery-powered field sensor

Build a low-power sensor node that wakes periodically, sends data, and sleeps to conserve battery.

What you'll build:

• ESP32 deep sleep cycle: wake every 5 minutes, connect to Wi-Fi and MQTT, publish readings, sleep
• Optimize connection time with static IP and BSSID caching
• Publish battery voltage alongside sensor data for remote monitoring
• Configure RosFit alerts for low battery (< 3.3V) and missed check-ins
• Achieve 6+ months on a 3000mAh LiPo battery

Hardware: ESP32, BME280 (temp/humidity/pressure), solar panel (optional), LiPo battery.

Key design decisions:

ESP32-CAM security camera

Stream JPEG snapshots from an ESP32-CAM module through RosFit to the dashboard camera feed widget.

What you'll build:

• ESP32-CAM capturing JPEG frames at configurable resolution
• Publish compressed frames to MQTT as base64-encoded payloads
• View the live feed in the dashboard camera widget
• Trigger snapshot capture via RosFit commands
• Configure motion detection alerts using frame differencing

Hardware: ESP32-CAM (AI-Thinker module), OV2640 camera.

Bandwidth considerations:

For MQTT, keep frame size under 256 KB (EMQX default max packet size). Use QVGA or VGA for reliable streaming.

BLE sensor mesh with RPi gateway

Build a mesh of BLE sensor nodes that report through a Raspberry Pi gateway to RosFit.

What you'll build:

• Multiple ESP32 nodes advertising BLE sensor data (temperature, motion, door state)
• Raspberry Pi running a BLE scanner and MQTT gateway
• Gateway registers as a gateway device type in RosFit
• Sub-device telemetry published under the gateway's topic namespace
• Dashboard shows the gateway with expandable sub-device list

Hardware: 3+ ESP32 BLE beacons, Raspberry Pi 4, USB BLE dongle (optional).

Architecture:

[ESP32 BLE #1] ──┐
[ESP32 BLE #2] ──┼── BLE ──▶ [Raspberry Pi Gateway] ── MQTT ──▶ [RosFit Cloud]
[ESP32 BLE #3] ──┘

The gateway aggregates BLE advertisements and publishes structured telemetry:

{
  "gateway_id": "gw-rpi-01",
  "sub_devices": [
    { "id": "ble-temp-01", "temperature": 22.5, "rssi": -45 },
    { "id": "ble-motion-02", "motion_detected": true, "rssi": -62 },
    { "id": "ble-door-03", "door_open": false, "rssi": -38 }
  ],
  "timestamp": "2026-03-30T10:05:00Z"
}

OTA firmware updates via RosFit

Push firmware updates to ESP32 devices over-the-air using the RosFit OTA system.

What you'll build:

• Compile ESP32 firmware with OTA partition scheme
• Upload firmware binary to RosFit via the API or CLI
• Create a deployment targeting individual devices or a group
• ESP32 firmware includes an OTA listener that downloads and applies updates
• Implement rollback on failed boot with a watchdog timer

Key steps:

1. Partition the ESP32 flash with two OTA slots (ota_0, ota_1)
2. Include the RosFit OTA client library in your firmware
3. Upload the binary: rosfit deploy firmware-v1.2.0.bin --version 1.2.0 --target greenhouse-sensors
4. The OTA service publishes the update URL to rosfit/{device_id}/ota/update
5. The device downloads, verifies the checksum, flashes to the inactive slot, and reboots
6. On successful boot, the device reports the new firmware version in its shadow state

# Upload and deploy firmware
rosfit deploy build/firmware.bin \
  --version 1.2.0 \
  --target greenhouse-sensors \
  --rollback-on-failure