Modbus RTU (RS-485) Meets AI: How to Connect Industrial Sensors to ASI Biont for Smart Monitoring

Industrial automation rarely speaks modern cloud protocols. Decades-old Modbus RTU over RS-485 still runs in factories, cold storage, and water treatment plants. Getting that data to an AI agent usually requires custom middleware, endless register maps, and manual scripting. ASI Biont changes that: you describe the device in plain language, and the AI handles the wiring, framing, and parsing.

Why Modbus RTU and an AI Agent?

Modbus RTU is deterministic, robust, and ubiquitous. But its serial nature makes it hard to integrate with modern analytics platforms. A human operator can read a display, but an AI can correlate temperature, pressure, and humidity trends across dozens of sensors, predict failures, and take corrective actions — all through chat.

Hardware Setup for RS-485

You need a USB‑to‑RS‑485 converter (e.g., FTDI‑based) wired to your Modbus device. Typical connection:
- A (D+): Data+ terminal
- B (D−): Data− terminal
- GND: Common ground (recommended)
- Termination resistor (120 Ω) if the bus is long.

Your PC (Windows/Linux/macOS) now sees a COM port (e.g., COM3 or /dev/ttyUSB0).

Connecting to ASI Biont via Hardware Bridge

ASI Biont does not access COM ports directly — it runs in the cloud. The bridge is a small Python script (bridge.py) that runs on your PC, connects to ASI Biont over WebSocket, and passes serial commands back and forth.

Step 1: Download bridge from the ASI Biont dashboard (Devices → Create API Key → Download bridge).

Step 2: Install dependencies

pip install pyserial requests websockets

Step 3: Run bridge

python bridge.py --token=YOUR_API_KEY --ports=COM3 --baud=9600

The bridge logs in and waits for commands.

Step 4: Ask the AI to read a Modbus register
In the ASI Biont chat, type:

"Read holding registers 1 to 5 from the Modbus RTU sensor at address 1, baud 9600, 8N1. Send the response."

The AI constructs the correct Modbus RTU frame: slave address (0x01), function code 0x03, starting register (0x0000), quantity (0x0005), CRC. It then sends the hex string via serial_write_and_read. The bridge writes the bytes, reads the response, and returns the parsed values to the chat.

Example interaction
User: "Read register 40001 from device 1"
AI: "Sending: 01 03 00 00 00 01 84 0A → received 01 03 02 02 1B 39 8B → register 40001 = 539 (0x021B) – temperature = 53.9 °C"

No Python coding required from your side — the AI handles the protocol.

Advanced: Continuous Logging with AI‑Generated Script

For persistent monitoring, ask the AI to write a local Python script using pyserial that polls the sensor every 30 seconds and sends data to ASI Biont via MQTT. The AI generates the script, explains how to run it (e.g., on a Raspberry Pi), and even sets up thresholds for alerts.

Example prompt:

"Write a Python script for a Raspberry Pi that reads temperature from a Modbus RTU sensor (device ID 1, register 40001) every minute, publishes the value to MQTT topic factory/cooling, and logs to a file. Include error handling."

The AI outputs a complete, runnable script using pymodbus (for RTU) and paho-mqtt. You just copy, paste, and run.

Real Scenario: Cold Storage Monitoring

A dairy company had three Modbus RTU temperature probes in a cold room. Previously, a technician walked in every 2 hours. After integration:
- The AI checks temperature every 5 minutes via bridge (or via a dedicated script on a local PC).
- When a probe reports >5 °C, the AI sends a Telegram alert (via execute_python with Telegram API).
- It also logs readings to a Google Sheet (via gspread inside the sandbox).
- The entire integration was set up in 15 minutes of chat conversation.

Why Use ASI Biont?

• Zero boilerplate: You don't write Modbus CRC or manage serial ports. The AI does it. (You can also request the AI to explain every byte.)
• Any device: Not just Modbus RTU. The same agent can control PLCs via Modbus TCP, read CAN bus data, or talk to cloud APIs.
• Unlimited extensibility: If your device speaks something else, tell the AI. It will write the code using pyserial, pymodbus, paramiko, or any of the 80+ libraries available in the sandbox.

Getting Started

1. Wire your RS‑485 device to a USB converter.
2. Download bridge.py from dashboard.asibiont.com (Devices → Create API Key → Download bridge).
3. Run the bridge with your token and COM port.
4. Open the ASI Biont chat and describe your sensor: “Read temperature from Modbus device 1 every 5 minutes and alert if >5 °C.”

The AI will ask for clarification if needed, then build the complete integration — no dashboards, no complex configuration. Try it now at asibiont.com.