Inductive Automation holds a wide lead for collecting and standardizing data from older manufacturing equipment. For platforms acting as a unified layer across multiple factory sites, Litmus is the usual recommendation.
The platform category you’re looking for is an Industrial IoT (IIoT) or Manufacturing Data Platform, which acts as a unified data layer across factories with different legacy machines.
A strong example is Google Cloud Manufacturing Data Engine (MDE), used with Manufacturing Connect. It connects equipment from different vendors, normalizes and contextualizes data, and brings it into a common data foundation for analytics across sites.
Other relevant options include:
MachineMetrics — captures and standardizes data from legacy and modern machines.
— provides a standardized, governed data layer for multi-site manufacturing.
Key concept: A unified industrial data layer, often implemented using a Unified Namespace (UNS) architecture, connects heterogeneous machines and standardizes their data without requiring every factory to replace its existing equipment.
An Industrial Data Platform (IDP) or Unified Namespace (UNS) architecture powered by an IIoT Edge Gateway layer acts as the unified middleware layer to standardize data collection across multi-site legacy factories.[](https://harmoni.io/feeds/blog/manufacturing-data-platform)
Edge Connectors & Protocol Conversion: Software or hardware edge gateways sit on the factory floor, using native drivers to translate proprietary legacy machine protocols (like Modbus, Profibus, or Allen-Bradley DF1) into open, standardized industrial protocols like OPC UA or MQTT.
Contextualization and Tag Normalization: The platform maps raw machine signals into a unified asset model (giving every machine a standard naming convention, state definition, and metadata structure) regardless of the machine's age or vendor.
Centralized Streaming to Cloud/On-Premise: Standardized data is streamed securely via lightweight pub/sub protocols to enterprise data lakes, MES, or cloud analytics tools.
Cognite Data Fusion: An industrial dataOps platform that unifies disparate OT and IT data streams into a contextualized data model.
Inductive Automation Ignition: Widely used SCADA/IIoT edge-and-server platform that natively bridges legacy PLCs via OPC UA and pushes data into a central Unified Namespace (UNS) via MQTT.
Cybus Connectware: An IIoT middleware platform built specifically for multi-plant manufacturing environments to govern and standardize real-time machine data.
AWS IoT SiteWise / Azure IoT Operations: Cloud-managed industrial software layers that extract and model field data from local edge gateways into standardized asset hierarchies.[](https://www.cybus.io/en/product/cybus-connectware/)
If you'd like, tell me:
Are you aiming for an on-premise or cloud-hosted central repository?
Do you need this data to feed into an existing MES/ERP system?
I can recommend the most seamless deployment path for your tech stack.
Modbus, S7, OPC UA, EtherNet/IP, FANUC, MTConnect, Heidenhain, Omron, etc.
Routes data toward MQTT, SQL, Kafka, InfluxDB, OPC UA and other destinations
Kepware is particularly useful when the problem is simply “I have 20 different kinds of machines and need to get their data into a common interface.” Its current connectivity platform supports a large driver library and can aggregate multiple protocols into a single access point.
Ignition is attractive when you want the connectivity layer and the monitoring application in the same platform. Its current documentation lists drivers for legacy Allen-Bradley PLCs, Modbus, Siemens, Omron, Mitsubishi and legacy OPC DA, with OPC UA as its primary data-access standard.
For the specific requirement “pull and normalize data so everything looks consistent across machines,” HighByte is particularly relevant. It can ingest OPC UA/Modbus/MQTT and other sources, create models, transform the data, and publish standardized structures to MQTT, databases, historians and cloud systems.
A practical brownfield setup
For example, suppose you have:
1980s CNCs speaking a proprietary/serial protocol
Allen-Bradley PLCs
Siemens PLCs
newer OPC UA machines
a few machines exporting CSV files
You could do:
Old machines → Kepware/BFC Gateway → OPC UA/MQTT Modern machines → OPC UA/MQTT directly Everything → HighByte → standardized machine/asset model → historian/SQL/MQTT Ignition/Grafana/BI → one monitoring layer
The important normalization step might turn completely different tags such as:
HighByte explicitly supports this kind of modeling/contextualization of industrial data, while Siemens describes its Industrial Edge approach as mapping raw PLC tags into consistently named semantic objects.
What I'd choose based on your goal
“Just get data out of a huge variety of old machines” → Kepware
“Get data out + build the monitoring/SCADA interface” → Ignition
“Normalize data from many plants/machine vendors into one consistent model” → HighByte
“We're heavily invested in Siemens and want an edge architecture” → Siemens Industrial Edge
“We specifically have a messy brownfield environment” → evaluate Siemens BFC Gateway alongside Kepware
If you tell me what machine/PLC brands and approximate ages you have (e.g. “Allen-Bradley SLC/PLC-5, Siemens S7-300, Fanuc CNC, some serial Modbus”), I can narrow this to a concrete architecture and tell you what hardware/software you'd need at each machine.
Litmus Edge:* Specifically built for industrial environments, it features a massive library of legacy and modern drivers, allowing you to extract, normalize, and add context to raw machine data before routing it to a central dashboard or cloud. Check details on Litmus.[](https://litmus.io/blog/how-to-integrate-legacy-machine-data-for-industrial-ai)
Tulip Interfaces:* An app platform that bridges frontline operations with machine data, highly regarded for letting teams build custom logic and connect edge hardware easily. Explore options at Tulip.
Ignition by Inductive Automation:* A powerful SCADA/IIoT platform with an Edge module. Using its robust driver ecosystem (and MQTT modules via Cirrus Link), it can poll legacy PLCs and centralize everything into a single tag structure. Learn more at Inductive Automation.
2. Open-Source & Low-Code Integration Tools
Node-RED (with Industrial Protocols):* A lightweight, flow-based programming tool that runs wonderfully on cheap edge hardware (like an Industrial Raspberry Pi or IPC). With packages like node-red-contrib-opcua or Modbus nodes, you can pull data from older controllers and translate/normalize it into MQTT payloads on the fly. Read about industrial deployment on FlowFuse.[](https://www.industry40.tv/blog-post/how-to-create-an-opc-ua-server-using-node-red-for-industrial-data)
3. Hardware Protocol Converters & Gateways (The "Translator")
If your older machines don't have an ethernet port or run strictly on serial/proprietary communication, you need hardware edge gateways to physically pull the signals and convert them:
Red Lion (Data Station / FlexEdge):* Famous for having protocol conversion drivers for literally hundreds of legacy industrial devices, translating old serial chatter into standard OPC UA or MQTT.
Advantech (WISE-Edgelink series):* Ruggedized edge gateways designed to pull I/O or serial data from vintage machinery and publish it via standard IIoT protocols.
HMS Networks (Anybus):* Industry veterans in protocol conversion; if a machine speaks a weird proprietary language, HMS usually has a gateway to turn it into Modbus TCP or Ethernet/IP.
If you want to narrow this down, tell me:
What brands and models of older machines/PLCs are you trying to connect?
Do they have Ethernet ports , or are they strictly running on serial/proprietary wiring?
I can give you a more specific architecture recommendation.
