Stop Chasing One Industrial Protocol. Layer Them Instead.

You're the controls engineer at a mid-sized food and beverage plant. You've got a line of Rockwell PLCs running EtherNet/IP, a few Siemens drives on PROFINET, and an aging Modbus RTU network for temperature sensors. Your boss just read an article about OPC UA and wants to know why you're not using it everywhere. You take a breath and explain that these protocols aren't rivals—they're a stack. That conversation is happening in plants everywhere, and it's the wrong conversation.

The protocol wars are a distraction

Every few years, someone declares a winner. Modbus is too old. PROFINET is too Siemens. EtherNet/IP is too Rockwell. OPC UA will replace everything. None of that is true. The reality is that industrial networking is layered, and the protocols that survive do so because they solve a specific problem at a specific layer. Modbus, PROFINET, and EtherNet/IP connect field devices and PLCs. OPC UA and MQTT carry data securely to SCADA, MES, and cloud systems (OPC Foundation). That's not a bug. It's the architecture.

Modbus isn't dead—it's the cockroach of industrial protocols

Modbus was developed in the late 1970s. It uses a master-slave register-based model and runs over both serial (RTU) and Ethernet (TCP). By any modern measure, it's primitive. No native security. No real-time guarantees. No device profiles. And yet the Modbus Organization calls it the world's most widely adopted industrial communication protocol (Modbus Organization). That's not nostalgia. That's utility. Modbus is simple enough to implement on a $2 microcontroller and well-documented enough that a new engineer can debug it with a laptop and a USB adapter. When you need to read a temperature from a sensor that was installed before you were born, Modbus is there. You don't rip it out. You wrap it.

PROFINET, EtherNet/IP, and EtherCAT are not interchangeable

The real-time Ethernet protocols each have a personality. PROFINET, developed by Siemens and PROFIBUS/PROFINET International, supports real-time (RT) and isochronous real-time (IRT) classes for motion control. EtherNet/IP runs the Common Industrial Protocol (CIP) over standard Ethernet and TCP/UDP, and it's especially common in North America on Rockwell/Allen-Bradley platforms. EtherCAT, originally from Beckhoff, was built for short cycle times (≤ 100 µs) and low jitter (≤ 1 µs). Those aren't marketing numbers. They reflect different design goals. If you're synchronizing a five-axis CNC, you care about EtherCAT's distributed clocks. If you're moving pallets in a warehouse, PROFINET RT is fine. Picking a protocol is not a religious decision. It's a mechanical one.

The strongest counter-argument: 'Just use OPC UA everywhere'

The most common pushback I hear is that OPC UA is the universal answer. It's vendor-neutral, platform-independent, and has built-in encryption and authentication. It's standardized as IEC 62541. It's the backbone for IIoT and IT/OT integration. All true. But OPC UA is not a fieldbus replacement. It's a client/server technology designed for vertical data movement—from machine to MES to ERP. You don't run a servo loop over OPC UA. You run it over EtherCAT or Sercos, then push aggregated data up through OPC UA. The same goes for MQTT. It's a lightweight publish/subscribe transport for constrained devices and IoT environments (ISO/IEC 20922 MQTT Standard). It's brilliant for sending sensor telemetry to a cloud broker. It's terrible for deterministic motion control. The counter-argument fails because it confuses layers with replacements.

Security doesn't live in the protocol—it lives in the architecture

Here's where the layered approach pays off. CISA warns that many legacy ICS devices use outdated operating systems and older protocols that lack encryption or authentication, and it identifies brownfield deployments—layering legacy infrastructure with modern OT systems—as a unique challenge (CISA Industrial Control Systems). That's exactly the situation in most plants. You can't rip out every Modbus device. But you can segment your network, put a firewall between the field layer and the SCADA layer, and terminate your OPC UA connections with certificates. ISA/IEC 62443 defines requirements for implementing and maintaining secure industrial automation and control systems, bridging the gap between operations technology and IT (ISA/IEC 62443). Use it. The protocol itself won't save you. The architecture will.

What to do on Monday

Stop looking for the one protocol to rule them all. Instead, audit your layers. Ask three questions:

• What's talking to my field devices? (Modbus, IO-Link, PROFINET, EtherNet/IP, EtherCAT, CANopen, Sercos)
• What's moving data between cells and SCADA? (OPC UA, MQTT)
• Where does security break down? (Probably at the boundary between old and new)

Then fill the gaps. If you have a mix of Rockwell and Siemens PLCs, use OPC UA as a normalization layer. If you have legacy Modbus sensors, put an IO-Link master in front of them and translate. If you're building a new line, pick the real-time protocol that matches your motion requirements, not the one your vendor's sales rep likes. The industrial automation market is projected to grow from $184.43 billion in 2025 to $326.48 billion by 2032 (Maximize Market Research). That growth will be built on heterogeneous networks, not monocultures.

The one thing to remember

Protocols are tools, not tribes. The engineer who insists on one protocol for everything is as wrong as the engineer who refuses to learn anything new. Layer them. Secure the boundaries. Move data up the stack. That's how real plants run.

Sources

• OPC Foundation - https://opcfoundation.org/
• Modbus Organization - https://www.modbus.org/
• CISA Industrial Control Systems - https://www.cisa.gov/topics/industrial-control-systems
• ISA/IEC 62443 - https://www.isa.org/standards-and-publications/isa-standards/isa-iec-62443-series-of-standards
• Maximize Market Research - https://www.maximizemarketresearch.com/