Integrating Modbus TCP and MQTT in Edge Computing Industrial Architectures

The Challenge of Connecting Traditional Factory Floors to the Modern Cloud

Modern industries face a constant dilemma. On one hand, there are old machines and robust control systems that have run nonstop for decades, yet were designed in complete isolation. On the other hand, managers and engineers demand real-time analytical dashboards, predictive artificial intelligence, and remote cloud access. In practice, this means we need to make systems that do not speak the same language communicate securely and efficiently.

The separation between the factory floor network and the corporate network has always been strict for operational safety and stability reasons. However, ignoring the data generated by machines limits the competitiveness of any industrial plant. This is where new engineering approaches step in, allowing us to extract useful insights from legacy equipment without jeopardizing the physical manufacturing process.

Understanding the Role of Modbus TCP in Industrial Networks

Modbus is one of the oldest and most widespread communication protocols in industrial automation. Created in the 1970s, it operates on a simple request-response logic, where a master device requests information from a slave device, such as a PLC (Programmable Logic Controller, a ruggedized industrial computer that commands machinery). Modbus TCP is the evolution of this protocol running over standard Ethernet networks and conventional network cables, replacing old serial connections.

The major advantage of Modbus TCP is its deterministic simplicity and the massive base of compatible equipment available on the market. In practice, any engineer can scan registers from a sensor using simple network tools. However, it has severe limitations for the modern internet: it requires the master to actively ask all the time (polling model), generating unnecessary traffic, and it was not natively designed to safely cross corporate routers or complex firewalls.

The Power of MQTT in Event-Driven Communication

To solve the limitations of traditional protocols in decentralized environments, the MQTT protocol has become the gold standard for industrial internet of things. MQTT stands for Message Queuing Telemetry Transport, a long name for an extremely lightweight protocol originally developed to monitor oil pipelines via satellite with poor connections. In practice, it works like a topic-based postal messaging system, where devices publish data only when a relevant change occurs, and interested servers subscribe to those topics to receive updates instantly.

Unlike Modbus, where the central system must ask for a temperature value every single second, an MQTT sensor sends data only when the temperature changes or at programmed intervals, saving massive amounts of network bandwidth and processing power. It uses a central intermediary called a broker (a message server), making it easier to connect hundreds of scattered devices without needing to open complex network ports for each individual machine.

The Concept of Edge Computing in Industrial Automation

Edge computing represents the decentralization of data processing. Instead of sending all raw data from thousands of sensors directly to a distant cloud or an overloaded central server, we place a compact, rugged computer (an edge gateway) physically close to the machines. In practice, this means intelligence and data filtering happen right inside the same warehouse where the product is being manufactured.

This physical proximity brings critical operational benefits. If the internet connection drops temporarily, the edge gateway keeps operating, storing data locally and preventing factory downtime. Furthermore, it drastically reduces the volume of data sent to the cloud by filtering out noise, calculating averages, and triggering immediate local alarms before any network delay can cause an accident or a spoiled production lot.

Practical Architecture for Modbus TCP and MQTT Integration

The convergence of these technologies typically occurs through an edge gateway running specialized software or custom scripts. The operational workflow works in well-defined steps. First, the gateway actively collects data from local PLCs using the Modbus TCP protocol in fast polling cycles. Next, internal software normalizes these raw numbers into standardized data structures, such as JSON or Sparkplug B. Finally, the gateway publishes this structured information via MQTT to a local or cloud broker.

Below is a practical example in Python running on an edge gateway that reads a Modbus TCP register and publishes the value via MQTT:

import time
import paho.mqtt.client as mqtt
from pymodbus.client import ModbusTcpClient

# Connection configuration
MODBUS_IP = '192.168.1.50'
MQTT_BROKER = 'broker.local'
TOPIC = 'factory/line1/temperature'

modbus_client = ModbusTcpClient(MODBUS_IP)
mqtt_client = mqtt.Client('EdgeGateway')
mqtt_client.connect(MQTT_BROKER, 1883)

while True:
    modbus_client.connect()
    result = modbus_client.read_holding_registers(address=100, count=1)
    if not result.isError():
        temperature = result.registers[0]
        mqtt_client.publish(TOPIC, str(temperature))
        print(f'Published data: {temperature}')
    modbus_client.close()
    time.sleep(5)

This script demonstrates the simplicity and effectiveness of translating a synchronous local protocol into an asynchronous event ready to be consumed by modern dashboards or artificial intelligence systems.

Security Considerations and Operational Best Practices

Integrating legacy industrial networks with the internet exposes factories to severe cybersecurity risks that did not exist before due to physical isolation. Therefore, implementing edge architectures requires the rigorous use of next-generation firewalls, network segmentation via industrial VLANs, and TLS encryption on MQTT connections. In practice, no field device should have direct internet access without passing through a strictly controlled industrial demilitarized zone.

Another critical point is managing credentials and digital certificates on gateways. Default factory passwords on PLCs and gateways must be changed immediately during commissioning. Continuous monitoring of edge traffic ensures that any anomalous behavior, such as unauthorized Modbus port scanning attempts, is detected and isolated before compromising the production line.

Final Considerations

The convergence of Modbus TCP and MQTT through edge computing represents a sustainable and financially viable path for industrial modernization. Instead of discarding entire machinery parks that still function perfectly, companies can adopt smart gateways to extract valuable data and fuel digital transformation. With proper architectural planning, rigorous cybersecurity, and resilient code, the factory floor gains agility without losing its historical reliability.