This project initially started with Concept 3, but was later upgraded to Concept 4 for improved scalability and real-world deployment. You can find visual diagrams for each concept in the concepts/ folder.
// TODO: User, please add a brief description of Concept 1 here.
// TODO: User, please add a brief description of Concept 2 here.
- ESP32 sensor nodes collect environmental data and send it directly to a central server or broker.
- Centralized approach, but less scalable for distributed, real-world deployments where each node is in a different home.
- Each ESP32 sensor node is paired with a Raspberry Pi gateway (Raspberry Pi 1) in a home.
- Sensor data is sent via UDP from ESP32 to the local Raspberry Pi 1.
- Raspberry Pi 1 uses DDNS and port forwarding to act as a gateway, publishing data to a remote MQTT broker.
- A second Raspberry Pi (Raspberry Pi 2) subscribes to the MQTT broker, writes data to a remote database, and hosts a web server for visualization.
- This distributed, gateway-based approach is more robust and scalable for real-world, multi-home deployments.
Why Concept 4?
- Concept 4 was chosen because it allows each ESP32 node to be deployed in a different home, with a local gateway (Raspberry Pi 1) handling connectivity and security via DDNS and port forwarding. This makes the system scalable, secure, and suitable for real-world use where direct access to each ESP32 from the internet is not feasible.
This repository implements a distributed IoT system for environmental sensing and data collection, based on Concept 4. The system is designed for real-world deployment, with each ESP32 sensor node and Raspberry Pi gateway located in different homes, using DDNS and port forwarding for remote connectivity. The architecture includes ESP32-C3 Super Mini boards, environmental sensors, two Raspberry Pi devices, a remote MQTT broker, and a remote database server.
-
ESP32 Sensor Nodes (PlatformIO projects in
platformio/):- Hardware: ESP32-C3 Super Mini boards
- Sensors: AHT20 (temperature/humidity), ENS160 (air quality/CO2/TVOC)
- Each ESP32 node is deployed in a separate home, collecting local environmental data.
- Sensor data is sent via UDP to a Raspberry Pi gateway (Raspberry Pi 1) in the same home.
-
Raspberry Pi 1 (raspberry1) – Home Gateway:
- Runs a UDP server to receive sensor data from the local ESP32 node (see
platformio/wifi_mqtt_test_concept4/udp_server_raspi_example.py). - Uses Dynamic DNS (DDNS) and port forwarding to be accessible from outside the home network.
- Acts as a gateway: receives UDP packets, parses them, and publishes the data to a remote MQTT broker.
- Example scripts and templates are in the
raspberry1/folder, including class examples for MQTT publishing.
- Runs a UDP server to receive sensor data from the local ESP32 node (see
-
Remote MQTT Broker:
- Hosted on a remote server (cloud or university server).
- Receives published sensor data from all Raspberry Pi 1 gateways (from different homes).
- Forwards data to all subscribers, including Raspberry Pi 2 and other clients.
-
Raspberry Pi 2 (raspberry2) – Remote Data Aggregator:
- Subscribes to the remote MQTT broker to receive sensor data from all homes.
- Saves incoming data to a remote database (e.g., InfluxDB) for long-term storage and analysis.
- Hosts a web server to visualize or provide access to the stored data.
- Example scripts for MQTT-to-DB and web server are in the
raspberry2/folder.
-
Remote Database Server:
- Stores all sensor data received via MQTT.
- Can be queried by the web server for visualization and analytics.
- DDNS & Port Forwarding: Each Raspberry Pi 1 uses a DDNS service (e.g., DuckDNS, No-IP) and router port forwarding to allow ESP32 nodes to send UDP packets from anywhere, and to enable remote management.
- Distributed Homes: Each ESP32 and Raspberry Pi 1 pair is installed in a different home, enabling a scalable, multi-location sensor network.
- Security: Ensure strong passwords and secure port forwarding on all home routers.
platformio/- ESP32 firmware projects (sensor, LED, WiFi/MQTT/UDP examples)raspberry1/- Gateway scripts for UDP-to-MQTT publishing, DDNS setup, and class MQTT examplesraspberry2/- Scripts for MQTT subscription, database writing, and web serverpymakr/- MicroPython projects for ESP32 (optional/legacy)concepts/- Project concept images and documentation
- ESP32 Nodes: Flash the appropriate firmware from
platformio/to your ESP32-C3 Super Mini boards. Connect AHT20 and ENS160 sensors as described in the project documentation. - Raspberry Pi 1: Set up DDNS and port forwarding. Run the UDP server and MQTT publisher scripts from
raspberry1/. - MQTT Broker: Set up a remote MQTT broker (e.g., Mosquitto, HiveMQ) accessible to all Raspberry Pis and the remote server.
- Raspberry Pi 2: Run the MQTT subscriber, database writer, and web server scripts from
raspberry2/. - Remote Database: Ensure the remote database server (e.g., InfluxDB) is running and accessible.
See the README files in each subfolder for detailed setup and usage instructions. Refer to the class example scripts in raspberry1/class_examples/ for MQTT publishing and other utilities.
- ESP32-C3 Super Mini boards with AHT20 and ENS160 sensors
- Two Raspberry Pi devices (or similar Linux SBCs)
- DDNS service and router port forwarding for each home gateway
- Remote MQTT broker
- Remote database server (e.g., InfluxDB)
- Python 3.x (for Raspberry Pi scripts)
- PlatformIO (for ESP32 firmware)
MIT