Ethernet-Based Controls: Connecting the Dairy Plant from Sensor to Cloud

Automation February 2025 · 7 min read

Twenty years ago, the state of the art in dairy plant automation was a PLC communicating with field devices over DeviceNet or Profibus, with a single SCADA workstation at the end of a serial RS-485 cable. Data moved slowly, networks were proprietary and vendor-locked, and the idea of a cloud-connected dairy plant was science fiction.

Today, industrial Ethernet has changed everything. Modern dairy plants are deploying the same TCP/IP networking fabric that runs the internet — running at gigabit speeds, using managed switches, VLANs, and firewalls — and it's enabling a level of connectivity and data accessibility that was simply impossible with legacy fieldbus.

From Fieldbus to Industrial Ethernet

The transition from proprietary fieldbus protocols to Ethernet-based alternatives has been underway for more than a decade. The major industrial Ethernet protocols in food and dairy are:

EtherNet/IP — The dominant protocol in North American food manufacturing. Developed by Rockwell Automation and now managed by ODVA. Runs standard TCP/IP and UDP, making it straightforward to integrate with IT infrastructure. Used extensively in Allen-Bradley PLC environments.
PROFINET — The Siemens-led industrial Ethernet standard. Widely used in European dairy equipment and becoming more common in North America through Siemens PLC deployments and German OEM machinery.
EtherCAT — High-speed deterministic protocol from Beckhoff, common in high-speed packaging and filling applications where sub-millisecond cycle times are required.
OPC-UA — Not a fieldbus, but a platform-neutral application-layer protocol that sits above the network layer. OPC-UA is the key to interoperability, allowing PLCs, historians, MES systems, and cloud platforms from different vendors to exchange structured data without custom integration work.

The IT/OT Convergence

When plant floor networks run standard Ethernet, the traditional hard boundary between operational technology (OT — PLCs, SCADA) and information technology (IT — ERP, databases, business analytics) begins to dissolve. This is both an opportunity and a challenge.

The opportunity: production data that used to be trapped in a PLC's memory can now flow — in near real time — into enterprise databases, business intelligence tools, and cloud platforms. A plant manager can see live production KPIs from a tablet at a trade show. A quality manager can pull up the exact temperature profile for any production batch from three years ago in seconds.

The challenge: opening OT networks to broader connectivity introduces cybersecurity risk. Industrial control systems that were never designed with security in mind can become exposed when connected to business networks. Best practices include network segmentation with DMZ layers between OT and IT, industrial firewalls, role-based access control, and regular patching schedules for HMI and SCADA software.

The Purdue Model: Most industrial cybersecurity guidance still references the Purdue Enterprise Reference Architecture, which defines five levels of automation — from field sensors (Level 0) to enterprise business systems (Level 4). Proper network segmentation means traffic between levels is controlled, logged, and filtered. In practice, many dairy plants are moving to a flattened three-zone model: field devices, plant operations (SCADA/MES), and enterprise (ERP/cloud).

Edge Computing in the Dairy Plant

Not all data needs to go to the cloud, and not all analytics can tolerate the latency of a round trip to a cloud server. Edge computing puts compute power — typically a hardened industrial PC or edge gateway device — at the plant level, where it can run local analytics, anomaly detection, and protocol translation without relying on a network connection.

Common edge computing use cases in dairy plants include:

Protocol conversion (Modbus → OPC-UA → cloud) for legacy instruments
Local batch record computation and storage as backup to cloud systems
Real-time SPC calculations that trigger PLC alarms without cloud dependency
Video analytics for product inspection and fill-level verification
Buffering historian data during WAN outages to prevent data loss

Cloud Connectivity Platforms

Several platforms have emerged specifically to connect industrial control systems to cloud data platforms. Ignition by Inductive Automation is widely used in North American food manufacturing — its MQTT/Sparkplug B architecture allows edge gateways to publish structured PLC data to cloud brokers with minimal configuration. AWS IoT Greengrass and Azure IoT Hub are also increasingly deployed in larger dairy operations connected to broader corporate cloud infrastructure.

The long-term vision — already being realized in the most advanced dairy operations — is a fully connected plant where every process variable is available to analytics tools in real time, historical data is stored indefinitely in low-cost cloud object storage, and machine learning models can be trained and deployed against live production data.

What to Prioritize When Modernizing

For most dairy plants in the middle of their automation journey, the highest-return investments are:

Replacing aging serial/fieldbus networks with managed Ethernet switches and EtherNet/IP or PROFINET where equipment is being replaced anyway
Deploying an OPC-UA server on PLCs to provide a vendor-neutral data access layer
Adding an edge gateway with historian capability to capture data that currently lives only in PLC memory
Establishing network segmentation between the plant floor and business network before expanding connectivity