LoRa Gateway

This directory contains the firmware, hardware wiring, and configuration details for the central gateway of the ForestProtector wildfire monitoring system. The gateway receives data from IoT sensor nodes via LoRa and performs processing based on a reinforcement learning algorithm. This algorithm enables the system to dynamically determine which area the thermal surveillance camera should focus on, optimizing zone monitoring effectiveness.

The intelligent processing is handled by a Jetson Nano, which receives data through a LilyGo T3S3 microcontroller equipped with LoRa connectivity. This setup ensures efficient, low-power communication while maintaining a modular and scalable architecture. The Jetson analyzes the input data and computes an optimal motion policy, generating commands that drive a camera positioning system to point toward the selected critical zone.

This system is designed to operate autonomously in forest environments, prioritizing high-risk areas based on the real-time sensor data collected by remote nodes.

Contents

1. Components
2. Schematic diagram
3. Files included
4. Physical design and assembly
5. Library setup
6. Firmware behavior
7. Field deployment considerations

1. Components

• LilyGo T3S3 V1.1 – ESP32-S3 microcontroller with integrated SX1262 LoRa transceiver
• Jetson Nano – Embedded computer with ARM Cortex-A57 CPU and Maxwell GPU, used for reinforcement learning-based decision making
• Pi Camera Module (Rev 1.3) – Camera connected to the Jetson Nano via CSI interface for thermal or visual monitoring
• SG90 Servo – Servo motor for pan or tilt control of the camera
• LoRa Antenna – External 868 MHz/915 MHz antenna for long-range communication
• 3D Printed Enclosure – Enclosure designed in PETG for weather resistance (UV and moisture protection)
• Power Supply – Power source compatible with Jetson Nano (e.g., 5V–4A via USB-C or 5.5 mm barrel connector)
• 2x 5V Fans – Used for active cooling of the Jetson Nano and airflow within the enclosure
• 20x32x7 mm Bearing – Provides smooth rotation support for the camera platform

2. Schematic diagram

*(to be completed)*

3. Files included

• gateway_code.ino: Arduino firmware for the LilyGo T3S3 microcontroller. It receives LoRa packets, parses the data sent by the IoT sensor nodes, and forwards it to the Jetson Nano via serial communication. It also receives commands from the Jetson to control the servo responsible for camera orientation.

• python/ (to be completed): Folder intended for the scripts that run on the Jetson Nano. These include the reinforcement learning algorithm, Q-table management, camera control, and communication with the microcontroller.

• cad_files/: Editable 3D design files created in SolidWorks.

  • The main part file for customizing the gateway enclosure is gateway_parts.SLDPRT.
  • gateway_with_electronics.SLDASM contains the full assembly with electronic components integrated in their intended positions.

• 3d_printing_files/: STL files ready for 3D printing.

  • This folder also includes gateway_all_parts.3mf, a pre-arranged layout with all parts correctly positioned and oriented for slicing and export.

4. Physical Design and Assembly

The physical design of the gateway is essential to ensure proper operation under various environmental conditions. We recommend housing the components in a weatherproof enclosure, such as an IP65-rated box, which will protect the electronics from dust, water, and other environmental factors. For this project, alternatively, the system can be placed under a roof or surface that shields it from direct sunlight and rain.

Recommended 3D printing settings:

• Material: PETG (preferred over PLA due to its superior outdoor durability)
• Layer height: 0.2 mm
• Infill: 20–30%
• Orientation: All STL files in the 3d_printing_files folder are pre-oriented for printing. Some parts may require supports, but all were designed to minimize or eliminate their use.

Assembly steps:

1. Print all parts available in the 3d_printing_files folder.
2. Connect all components to their respective cables: Jetson Nano to its power supply, T3S3 to its power cable, and the camera to the CSI cable.
3. Insert the servo_gear_camera_drive.stl part onto the already mounted SG90 servo shaft, securing it with the screws provided with the motor.
4. Mount the two 5V fans to the bottom of the gateway_enclosure.stl part and secure them using appropriate screws.
5. Place the T3S3 board, Jetson Nano, and SG90 servo motor in their designated slots inside gateway_enclosure.stl. Secure the Jetson and the motor using one screw each.
6. To secure the T3S3, insert the LoRa antenna through the top opening of gateway_enclosure.stl, and fix it using the matching nuts and adjustment pieces.
7. For the camera assembly: first, attach the camera module to camera_module_cover.stl using the appropriate screws. Insert a bearing into camera_gear_connector.stl, then mount camera_cable_pillar.stl into the bearing. Connect all parts with screws. The bearing allows smooth movement of the assembly.
8. Next, join camera_module_enclosure.stl to the previously assembled camera module using the same screws. Route the camera cable through the provided slot and secure camera_module_enclosure.stl to camera_module_cover.stl, completing the pan-tilt structure.
9. Insert the fully assembled camera module into gateway_enclosure.stl, ensuring that the gears mesh properly. Secure the module with two screws; the height will auto-align.
10. Additional recommendation: Ensure that the servo motor is set to its initial position before meshing the drive gear and camera gear. This step ensures proper alignment of movement and helps protect the CSI cable from torsion or damage.
11. Finally, close the enclosure by screwing on the gateway_enclosure_cover.stl part, using it as a lid. Make sure to leave the Jetson Nano's power cable routed through the dedicated notch or opening to exit the enclosure safely.

Assembly reference:

5. Library Setup

To compile the gateway firmware and run the scripts, you need the following:

Arduino IDE (for LilyGo T3S3)

• Board support: Install "ESP32 by Espressif Systems" by adding the following URL in Preferences:
  https://raw.githubusercontent.com/espressif/arduino-esp32/gh-pages/package_esp32_index.json

• Required libraries:

  • LoRa by Sandeep Mistry

• Development environment configuration:
  When using the LilyGo T3S3 board, you must manually adjust the board settings in the Arduino IDE to ensure proper operation. Use the following recommended configuration:

Option	Value
Board	ESP32S3 Dev Module
Port	Your board’s port
USB CDC On Boot	Enable
CPU Frequency	240MHz (WiFi)
Core Debug Level	None
USB DFU On Boot	Disable
Erase All Flash Before Upload	Disable
Events Run On	Core1
Flash Mode	QIO 80MHz
Flash Size	4MB (32Mb)
Arduino Runs On	Core1
USB Firmware MSC On Boot	Disable
Partition Scheme	Huge APP (3MB No OTA / 1MB SPIFFS)
PSRAM	QSPI PSRAM
Upload Mode	UART0 / Hardware CDC
Upload Speed	921600
USB Mode	CDC and JTAG
Programmer	Esptool

Also, make sure to uncomment the appropriate board in the utilities.h file of your sketch according to the model you're using. Otherwise, compilation will fail.

For more details on configuring this board, refer to the official LilyGo documentation:
LilyGo LoRa Series GitHub Repository

(for Jeston Nano)

(to be completed)

6. Firmware Behavior

The gateway system consists of two main components:

LilyGo T3S3 (LoRa Gateway)

This microcontroller acts as the central receiver for environmental monitoring data sent from remote sensor nodes. Its operation can be summarized as follows:

1. LoRa Module Initialization
   In the setup(), the SPI pins for the LoRa module are manually defined, and the transceiver is initialized at 915 MHz. If the initialization fails, the system halts.

2. Data Reception
   Inside the loop(), the device continuously listens for incoming LoRa packets using LoRa.parsePacket(). When a packet is received:

   • The full message is read into a String.
   • The processReceivedData() function is called to interpret its contents.

3. Received Data Processing
   The expected message format is:
   <NODE_ID>:<mq2>,<rain>,<temperature>,<humidity>
   Example: 2:1234,0,55.3,48.1
   Where:

   • NODE_ID identifies the transmitting node (1, 2, or 3).
   • mq2 is the raw smoke sensor value (mapped to a 0–100 scale).
   • rain is the raw rain sensor value.
   • temperature and humidity are in °C and %RH respectively.

   The function:

   • Associates the data to the corresponding node (node1, node2, or node3).
   • Applies binary thresholds: temperature > 50°C and humidity > 45% are set to 1; otherwise, 0.
   • Logs the last update time for the node.

4. Periodic Display
   Every 5 seconds, the displayNodeData() function prints the latest data for all nodes to the Serial Monitor. If a node has not updated its data in more than 10 seconds, it reports "No recent data received."

Jetson Nano (AI Agent and Decision Unit)

(to be completed)

This architecture enables the gateway to function as an intelligent hub, receiving, interpreting, and routing data to the AI-based decision unit.

7. Field Deployment Considerations

When deploying the gateway in real-world scenarios, take the following aspects into account:

Antenna Placement

• Position the LoRa antenna in a high, unobstructed location with line-of-sight to the sensor nodes.
• Avoid major obstacles (walls, dense trees, metallic structures) that may degrade the signal.
• If needed, use external high-gain antennas to increase communication range.

Power Supply

• Use a stable power source, ideally a USB-C power bank with reliable output.
• Alternatively, consider a solar panel + rechargeable battery solution for long-term remote deployments.

Environmental Protection

• The gateway must be housed in a weather-resistant enclosure:
  • Waterproof (IP65 or higher).
  • UV-resistant if exposed to direct sunlight.
• Properly seal all joints and leave a secure path for the power cable without exposing the system to water ingress.

Cooling

• Although the Jetson Nano includes fans, make sure they are functioning properly, especially in warm environments or during extended high-load operation.

Maintenance

• Perform regular checks on:
  • Servo alignment and movement.
  • Cable and connector integrity.
  • Camera focus and positioning.
• Ensure that communication with sensor nodes remains stable over time.

This modular gateway design allows future expansion, such as multiple LoRa channels, cloud service integration, or additional modules for intelligent decision-making and alerting.