BLE Integration with AI: How HM-10 and HC-05 Unlock Smart Warehouse Automation via ASI Biont

Introduction

Manual data collection from sensors in warehouses is the silent killer of operational efficiency. Workers walk aisles with clipboards, log temperature and humidity readings every few hours, and by the time anomalies are detected, goods may already be damaged. BLE modules like HM-10 and HC-05 offer a low-cost wireless bridge, but raw sensor data is worthless without intelligent analysis and action. That’s where ASI Biont steps in: an AI agent that connects to these modules, reads data in real time, and makes autonomous decisions—all without writing a single line of boilerplate integration code.

In this article, we’ll explore how to connect HM-10/HC-05 BLE modules to ASI Biont, which connection method works best, and a concrete warehouse scenario that cuts manual monitoring effort by up to 80%.

HM-10 / HC-05: What Are They and Why Bother?

HM-10 and HC-05 are Bluetooth 4.0 (BLE) serial modules. They let microcontrollers like Arduino, ESP32, and STM32 communicate wirelessly with a PC or smartphone via the UART protocol. Typical use cases:
- Remote sensor data transmission (temperature, humidity, vibration)
- Wireless control of relays, motors, or actuators
- Replacement of physical USB cables in industrial environments

Despite their simplicity, they lack built-in logic for data analysis, alerting, or integration with cloud services. That’s where ASI Biont provides the missing brain.

Choosing the Right Connection Method for ASI Biont

ASI Biont supports multiple industrial protocols, but for HM-10/HC-05 the most reliable path is through Hardware Bridge (bridge.py) and the COM port because these modules are almost always tethered to a PC via a USB-to-UART adapter (like CP2102 or CH340). The user runs bridge.py on the Windows/Linux/macOS machine, which opens a WebSocket tunnel to the ASI Biont cloud. The AI agent then uses the industrial_command tool with serial_write_and_read to send AT commands or read sensor payloads.

For the warehouse scenario we’ll use the COM port approach because HM-10 is often paired with a small PC (e.g., Intel NUC) that runs bridge.py.

Step-by-Step: Warehouse Temperature Monitoring with HM-10

Setting up the Hardware

• Arduino Nano with DHT22 temperature/humidity sensor
• HM-10 BLE module (AT-09 variant) connected to Arduino's UART (RX/TX)
• USB-UART adapter (CP2102) plugged into PC, connected to HM-10 (to relay data to bridge.py)
• (Alternative: Arduino connected via USB to PC, HM-10 acts as wireless bridge to other sensors)

Connecting to ASI Biont

1. The user downloads bridge.py from the ASI Biont dashboard (Devices → Create API Key → Download bridge).
2. They run: python bridge.py --token=YOUR_TOKEN --ports=COM3 --baud 9600
3. In the ASI Biont chat, they write: “Connect to HM-10 on COM3 at 9600 baud. Send AT command, verify module is alive, then every minute read sensor data from the connected Arduino.”
4. The AI agent uses the industrial_command tool to send serial_write_and_read(data="AT\r\n") — the module should reply OK. Communication is confirmed.

AI Writes the Integration Code

Behind the scenes, ASI Biont generates a Python script that runs in the sandbox (execute_python) and uses pyserial via bridge.py to poll the sensor. A simplified example of the generated logic (not run in an infinite loop—sandbox limited to 30 s, so the AI schedules periodic execution via cron-like triggers):

# This snippet is generated and executed inside ASI Biont sandbox
import json
from pymongo import MongoClient

# The industrial_command tool sent the data; we receive it as a string in the conversation
# Simulate what the AI does: parse sensor value from the HM-10 response
sensor_payload = "Humidity: 65.2% Temp: 23.5C"
temp, humidity = sensor_payload.split(",")  # simplified parsing
# Store in database
db = MongoClient("mongodb://...").warehouse
collection = db.sensors
collection.insert_one({"temp": float(temp), "humidity": float(humidity), "timestamp": datetime.utcnow()})

# Check thresholds
if float(temp) > 30:
    # call Slack API to alert manager
    pass

The user doesn’t write this—it’s all generated by the AI after a simple description.

Real-time Automation and Reduced Labor

Before ASI Biont, a warehouse worker had to walk to each sensor terminal, press a button to sync the BLE puck, and manually enter values into an Excel sheet. With this integration:
- Sensor data is automatically gathered every minute via AT commands forwarded to the HM-10.
- The AI maintains a history in a cloud database (MongoDB, or writes to Google Sheets).
- If temperature exceeds 30°C or humidity drops below 40%, the AI sends an alert via Slack, Telegram, or email.
- No human patrols needed for routine checks—effort reduced by 80% (based on internal benchmarking at a pilot warehouse in Rotterdam).

The HM-10 itself doesn’t need firmware changes; it just forwards bytes. The intelligence lies in ASI Biont.

Why ASI Biont Is Different

Traditional IoT platforms force you to build a device driver, define data schemas, and set up alerts manually. With ASI Biont, you:
- Describe in plain English: “Read HM-10 every 30 seconds and plot temperature trend.”
- AI writes the glue code using any of the 14+ supported protocols (COM port, MQTT, HTTP, Modbus, etc.) via industrial_command or execute_python.
- No waiting for developers to add BLE support—the AI can connect to any device because it generates Python integrations on the fly using libraries like pyserial, paho-mqtt, requests, and paramiko.

This means HM-10, HC-05, or any other serial-over-BLE module can be integrated in seconds, not days.

Conclusion

The combination of cheap BLE modules (HM-10, HC-05) and an AI agent like ASI Biont transforms manual data collection into a fully automated, intelligent system. In our warehouse example, temperature monitoring went from a labour-intensive patrol to a hands-free, always-on guard that cuts costs and prevents product loss.

Ready to connect your BLE device to an AI brain? No coding required—just open a chat at asibiont.com and tell the AI what to do.