LEVEL 1
Small prototype
- Gateway or script
- MQTT
- Node-RED
- Database and Grafana
For testing, internal demonstrations, and validating an initial data model.
Not production-critical infrastructure yet.
Real Industry 4.0
Industry 4.0 should start with one useful signal.
Not with another platform, a multi-year transformation program, or 800 PLC tags nobody will use.
Real Industry 4.0 is a practical approach to moving data from existing machines into systems people actually use—safely, clearly, and one step at a time.
Practical industrial data engineering for existing factories.
A practical position
Industry 4.0 is not a product.
Buying a platform does not automatically make an existing factory digital. A dashboard on top of unreliable data does not improve the data. And an MQTT broker alone does not solve an operational problem.
The useful starting point is smaller: one concrete question, a few trustworthy signals, and a data path someone can understand and operate.
“The interesting part begins when one value reliably leaves the machine and someone actually uses it.”
The One Signal framework
The practical path starts small.
Seven steps from an operational problem to a data path that still works after the prototype.
- 01
The question
Which operational problem should the data answer?
- Why did the line stop?
- How many parts were produced?
- Is the gateway still online?
- Is cycle time increasing?
- Which alarm occurred before downtime?
- 02
The signal
Which small set of values is actually required?
- 03
The source
Does the value already exist in a PLC, drive, meter, SCADA system, or database?
- 04
The data path
How does the data safely leave the machine or plant network?
- 05
The data model
How are name, unit, timestamp, quality, and source represented?
- 06
The use
Who uses the data, and which decision does it improve?
- 07
The operation
How are failures, buffering, backups, versions, and ownership handled?
From signal to decision
An example: Why does the conveyor stop?
Operational question
Why does Conveyor 3 stop during normal production?
Minimum useful signals
- Motor running
- Fault active
- Safety circuit active
- Upstream material available
- Downstream blocked
- Product counter
- PLC timestamp
Source
Existing Siemens PLC
Data path
- Siemens PLCread-only
- Industrial connector
- MQTT or database
- Historian and dashboard
Useful result
A stop timeline showing whether the conveyor stopped because of a local fault, upstream starvation, downstream blockage, or safety interruption.
What is deliberately excluded
- Every available PLC tag
- AI prediction
- Plant-wide architecture
- New sensors before existing data is checked
- A new MES
- A large cloud platform
The first success is not the dashboard. It is a data path whose values are understood and trusted.
Three levels
As large as necessary. No larger.
The architecture should follow operational responsibility and use-case scope—not a product category.
LEVEL 2
Local production setup
- PLC / OPC UA / Modbus
- Local integration layer
- Historian / MQTT / database
- Dashboard / MES / reports
For permanently operated data flows within one plant.
LEVEL 3
Multi-site data architecture
- Plant sources
- Local plant data layer
- Contextualized events and values
- Enterprise systems
For multiple sites, governed data models, and central applications.
Not every project needs a Unified Namespace, a data lake, or an enterprise platform.
A practical technology guide
Tools, not religions.
The right technology depends on the source, network boundaries, data shape, ownership, and operating model.
Direct PLC access
- Useful when
- A controlled, read-only connection to an existing PLC is the simplest available source.
- Be careful when
- Multiple consumers begin connecting directly or operational restrictions prohibit it.
OPC UA
- Useful when
- A structured and maintained OPC UA server already exposes the required data.
- Be careful when
- The server exposes only a limited subset or weak context.
Modbus TCP
- Useful when
- Existing devices expose simple process values through a well-documented register map.
- Be careful when
- Data types, endianness, scaling, or network protection are not clearly defined.
MQTT
- Useful when
- Values or events must be distributed to several consumers.
- Be careful when
- The topic structure and payload model have not been agreed.
REST
- Useful when
- Business applications need bounded data or functions on demand.
- Be careful when
- Polling, latency, or unreliable connections are underestimated in a real-time data path.
Database integration
- Useful when
- Existing operational records must support reports or clearly defined applications.
- Be careful when
- Consumers become coupled to internal tables or write access and data ownership are unclear.
Node-RED
- Useful when
- A small team needs to prototype, transform, and route data quickly.
- Be careful when
- Undocumented flows quietly become business-critical production software.
Grafana
- Useful when
- Teams need flexible technical dashboards over time-series or database data.
- Be careful when
- A dashboard is used as a substitute for a reliable data model or operational application.
From the field
What regularly goes wrong
- 01
Collect everything because storage is cheap.
- 02
Give every application direct PLC access.
- 03
Start with AI before timestamps are correct.
- 04
Use Node-RED without versioning or ownership.
- 05
Build one large dashboard nobody opens.
- 06
Store values without units or quality information.
- 07
Treat a successful prototype as production-ready.
- 08
Buy a platform before agreeing on the actual question.
- 09
Turn temporary edge hardware into critical infrastructure.
- 10
Assume more real-time data is always more useful.
“A dashboard on top of bad data is just a nicer way to be wrong.”
Free and practical
Practical resources instead of more slides
Calculators, a test environment, and an upcoming guide for the path from one signal to a dependable data flow.
Free calculators and converters
Tools for PLC analog values, Modbus registers, three-phase power, voltage drop, and drive engineering.
Open toolsFree Industrial Data Stack
MQTT, Node-RED, time-series storage, and Grafana in a prepared test environment.
Explore Sync StackComing soon
Brownfield Data Playbook
A practical guide from the first machine signal to a reliably operated data path.
When the prototype stays
When the first signal becomes a permanent system
Sync Motion supports companies with brownfield machine connectivity, local industrial data paths, and production-ready OT/IT integration.
Continue reading
Technical foundations and practical guidance
The useful starting point
Do not start with the platform. Start with the question.
When one value reliably leaves the machine, is understood, and improves a real decision, more has been achieved than in many large transformation programs.
- Question
- Signal
- Source
- Data path
- Useful outcome