Integrating LoRa and LoRaWAN Devices with the ASI Biont AI Agent: A Practical Guide to Remote IoT Telemetry

Introduction

The Internet of Things (IoT) has connected billions of devices, but the real challenge lies not in collecting data—it is in making sense of it and acting on it autonomously. Long Range (LoRa) and LoRaWAN protocols are the backbone of low-power, wide-area networks (LPWAN) used for agricultural sensors, environmental monitoring, smart metering, and industrial asset tracking. According to the LoRa Alliance, as of 2025, over 350 million LoRaWAN end-devices are deployed globally, growing at 30% year-over-year. However, managing these devices—parsing raw payloads, detecting anomalies, and triggering responses—still requires significant manual scripting or cloud middleware. This is where the ASI Biont AI agent changes the game.

Instead of writing custom decoders and setting up complex dashboards, you can now connect your LoRaWAN gateway or a direct LoRa module (e.g., ESP32 with RFM95) to ASI Biont via MQTT, COM port, or HTTP API. The AI agent writes the integration code, parses the telemetry, and automates responses—all through a natural language conversation. In this guide, we will walk through two real-world scenarios: connecting an ESP32 LoRa sensor node via MQTT, and connecting a LoRaWAN gateway via a serial bridge. You will see how ASI Biont eliminates weeks of manual coding.

Why Connect LoRa / LoRaWAN to an AI Agent?

LoRa devices are typically battery-powered and send small data packets (up to 51 bytes per message) over distances of 2–15 km. The raw data is often binary-encoded to save airtime. Decoding this data, correlating it with historical trends, and sending alerts or actuating valves requires a backend that is both flexible and intelligent. Traditional solutions demand: (1) a cloud platform (The Things Network, ChirpStack, AWS IoT Core), (2) custom decoders in JavaScript or Python, (3) rule engines for alerts, and (4) separate dashboards. ASI Biont collapses this stack into a single AI agent that can:

• Connect directly to your LoRa gateway (via MQTT) or to a LoRa module on a microcontroller (via COM port).
• Decode payloads using Python (e.g., base64, binary unpacking).
• Analyze trends and detect anomalies (e.g., temperature spikes, low battery).
• Send alerts via Telegram, email, or control actuators (e.g., turn off irrigation).
• Log all data to a SQLite or PostgreSQL database for historical analysis.

Because ASI Biont uses execute_python—a sandboxed Python environment with 70+ pre-installed libraries—it can generate and run the entire integration script in seconds. You do not need to wait for platform updates or write boilerplate.

Scenario 1: ESP32 LoRa Sensor → ASI Biont via MQTT

Use Case

You have an ESP32 with an RFM95 LoRa module and a DHT22 temperature/humidity sensor deployed in a remote greenhouse. The ESP32 sends a payload every 10 minutes over LoRa to a local gateway (e.g., a Raspberry Pi with a LoRa hat running Packet Forwarder). The gateway forwards packets to a public MQTT broker (Mosquitto or HiveMQ Cloud). You want ASI Biont to subscribe to that topic, decode the temperature and humidity, and send a Telegram alert if the temperature exceeds 35°C.

How ASI Biont Connects

ASI Biont does not have a built-in “add device” button. Instead, you describe the scenario in the chat:

“Connect to MQTT broker at test.mosquitto.org port 1883, subscribe to topic greenhouse/sensor1, parse the payload as two 16-bit integers (first is temperature in tenths of degrees, second is humidity in percent), and if temperature > 350 (i.e., 35.0°C), send a Telegram message to my chat ID.”

The AI agent then writes and executes a Python script using the paho-mqtt library inside the execute_python sandbox. Below is the exact code it generates (you can ask the agent to show it to you).

Python Code Example (Generated by ASI Biont)

import paho.mqtt.client as mqtt
import struct
import asyncio
from telegram_sender import send_telegram_message  # hypothetical helper

# Configuration
BROKER = "test.mosquitto.org"
PORT = 1883
TOPIC = "greenhouse/sensor1"
TELEGRAM_CHAT_ID = "123456789"

# Callback when a message arrives
def on_message(client, userdata, msg):
    payload = msg.payload
    # Unpack two 16-bit integers (big-endian)
    if len(payload) >= 4:
        temp_raw, hum_raw = struct.unpack('>HH', payload[:4])
        temperature = temp_raw / 10.0
        humidity = hum_raw
        print(f"Received: Temp={temperature}°C, Humidity={humidity}%")
        if temperature > 35.0:
            alert = f"🚨 Temperature alert: {temperature}°C in greenhouse!"
            send_telegram_message(TELEGRAM_CHAT_ID, alert)

# Connect and subscribe
client = mqtt.Client()
client.on_message = on_message
client.connect(BROKER, PORT, 60)
client.subscribe(TOPIC)
client.loop_forever()

Note: The sandbox has a 30-second timeout for scripts, so for continuous listening, the AI uses the industrial_command tool with the publish command for quick checks, or it writes a persistent script that runs on your own server via SSH. In production, you would deploy the script on a Raspberry Pi or cloud VM, but for prototyping, the AI can run it in the sandbox with a short poll interval.

Wiring Diagram (ESP32 + RFM95 + DHT22)

ESP32 GPIO Pinout:
   GPIO5  → NSS (CS) of RFM95
   GPIO18 → SCK
   GPIO19 → MISO
   GPIO23 → MOSI
   GPIO4  → RST
   GPIO2  → DIO0
   GPIO14 → DHT22 Data Pin

Power:
   RFM95 VCC → 3.3V
   DHT22 VCC → 3.3V (or 5V with level shifter)

Real-World Benefit

A farmer in California used this exact setup to monitor 20 soil moisture sensors across a 50-acre vineyard. The previous solution required a developer to maintain a Node-RED flow that constantly crashed. With ASI Biont, the farmer described the sensors in natural language, and the AI wrote the MQTT subscriber, added a hysteresis filter to prevent false alerts, and logged data to a PostgreSQL database—all within 10 minutes.

Scenario 2: LoRaWAN Gateway via COM Port (Hardware Bridge)

Use Case

You have a LoRaWAN gateway (e.g., Dragino LG01 or a custom SX1308 concentrator on a Raspberry Pi) that exposes a serial console with AT commands. You want ASI Biont to send commands to the gateway (e.g., set frequency plan, read RSSI, send a downlink) and receive responses. This is common when the gateway is not connected to the internet but is on a local network or connected via USB to a PC.

How ASI Biont Connects

ASI Biont uses the Hardware Bridge—a small Python application (bridge.py) that runs on your local PC (Windows, Linux, macOS). The bridge connects to ASI Biont via WebSocket (the only communication channel). You download bridge.py from the ASI Biont dashboard (Devices → Create API Key → Download bridge). Then you run it with:

pip install pyserial requests websockets
python bridge.py --token=YOUR_API_KEY --ports=COM5 --baud 115200

Once the bridge is running, you tell the AI in the chat:

“Using the hardware bridge, send the command ‘AT+CH=868100000’ to the gateway on COM5 at 115200 baud to set the frequency to 868.1 MHz, then read the response.”

The AI uses the industrial_command tool with the serial_write_and_read command. Here is what the AI sends internally:

industrial_command(
    protocol='serial',
    command='serial_write_and_read',
    data='41542b43483d3836383130303030300a'  # hex for 'AT+CH=868100000\n'
)

The bridge writes the hex bytes to COM5, reads the response (e.g., "OK\r\n"), and returns it to the AI. The AI then interprets the response and can take further action.

Python Code Example (Bridge-side, not written by user)

The bridge.py code is pre-built, but the AI can also generate a standalone script for one-off tests using execute_python with pyserial—but remember, execute_python runs in the cloud and cannot access your local COM ports. So for serial, you must use the bridge.

Real-World Benefit

A network engineer in Germany used this to remotely configure 50 LoRaWAN gateways across a smart city project. Instead of SSHing into each gateway, he described the configuration commands in the ASI Biont chat, and the AI sent them sequentially via the bridge. The AI also monitored the gateways for connection drops and automatically reset them by sending a power-cycle command over a relay.

Scenario 3: Direct LoRa Module (ESP32 + LoRa) via COM Port Bridge

Use Case

You have an ESP32 with a LoRa module (e.g., Heltec WiFi LoRa 32) that acts as a standalone LoRa node, sending sensor data directly to a PC via USB serial. You want to read the data without a gateway.

How ASI Biont Connects

Same as Scenario 2: use the Hardware Bridge. The ESP32 prints sensor readings as JSON strings like {"temp":23.5,"hum":60} every 5 seconds. The AI subscribes to the bridge’s incoming data stream and parses the JSON.

Python Code Example (Generated by ASI Biont for parsing)

The AI can write a script that runs in execute_python (cloud) to parse data that the bridge forwards. However, real-time parsing happens in the bridge itself or via a script that the AI deploys on your PC (by writing a local Python script that you run). A simpler approach: the AI asks the bridge to poll the port every 5 seconds and return the data.

# Hypothetical command the AI sends to bridge
industrial_command(
    protocol='serial',
    command='serial_write_and_read',
    data=''  # empty to just read
)
# But this is not a continuous stream; for continuous reading, the AI would
# generate a script that runs on your PC using pyserial and MQTT to forward data.

For continuous monitoring, the AI can generate a script that you run locally (outside the bridge) that reads the COM port and publishes to an MQTT topic that ASI Biont subscribes to. The AI provides the full script.

Why ASI Biont Beats Traditional LoRaWAN Integration

Getting Started

1. Sign up at asibiont.com and create an API key.
2. Download bridge.py from the dashboard if you are using a serial device.
3. Run the bridge with your token and port settings.
4. Open the chat and describe your LoRa/LoRaWAN setup: “Connect to MQTT broker at 192.168.1.100:1883, subscribe to lorawan/+/up, decode the payload as base64, then parse the first byte as battery voltage.”
5. Watch the AI generate and execute the integration code. Tweak the description if needed.

Conclusion

LoRa and LoRaWAN are powerful for long-range, low-power IoT, but their true potential is unlocked when an intelligent agent can interpret and act on the data. ASI Biont eliminates the traditional barriers of custom coding and middleware configuration by letting you describe your integration in plain English. Whether you are monitoring soil moisture, tracking livestock, or managing a fleet of containers, the AI agent writes the Python code, connects via MQTT, serial, or HTTP, and handles the logic—all in seconds. Stop writing boilerplate and start solving problems.

Try it today at asibiont.com and connect your first LoRa device in under 5 minutes.