Industrial IoT Monitoring Dashboards with ThingsBoard
Custom ThingsBoard dashboards for device telemetry, industrial sensor measurements, equipment status, operational trends and alerts.
Project status: In development
Most plants do not lack data. It is locked inside PLCs, switchboards and shift logbooks. The production manager knows the line missed its target today, but not which short stops caused it or which machine is quietly driving up the energy bill.
As part of this smart factory service, I collect the data your machines already produce, turn it into KPIs that operators and managers understand, and connect it to MES or ERP where that adds value. The emphasis is on existing equipment and a staged rollout, so production keeps running while the system grows.
Industry challenges
Mixed machine generations
A single line often combines PLCs from different eras, proprietary controllers and machines with no communication interface at all.
Invisible micro-stops
Stops of a few minutes and reduced speed are rarely recorded, yet together they consume a meaningful share of capacity.
Energy costs without attribution
The utility bill covers the whole site, leaving open which line, compressor or furnace is responsible for most of it.
The OT/IT gap
Shop-floor and office networks are separate and often incompatible; linking them without a security concept creates real risk.
Technical solutions
Machine data acquisition
Reading states, counters and alarms from PLCs via OPC UA, Modbus or vendor drivers, without changing control logic.
OEE-style production KPIs
Availability, performance and quality derived from real machine data, with stop reasons classified together with operators.
Energy monitoring
Energy meters or current transformers on major consumers, with consumption shown per line, shift and product.
Predictive maintenance basics
Vibration, temperature and current trends on critical assets, with thresholds that alert maintenance before a breakdown.
MES and ERP integration
Production counts, downtime and material consumption passed to planning systems through APIs or an intermediate database.
Network architecture and OT security
Segmentation between production and office networks, with an edge gateway as the controlled point of data exchange.
From first machine to connected plant
Production line digitalization is not a one-off installation. It starts with one line or one bottleneck and expands based on what the data shows. A typical sequence:
- Shop-floor assessment: inventory machines, controllers, available protocols, network topology and how production information flows today.
- Goals and KPIs: agree with production management which questions the data must answer, for example the main causes of downtime or the energy used per product.
- Data architecture: machine connectivity, edge gateway, MQTT transport, data platform and dashboards, built on one consistent data model for all machines.
- Pilot line: connect the machines, validate the numbers with operators and refine KPI definitions until they reflect reality.
- Scale and integrate: add further lines, connect to MES or ERP and train the users who will rely on the system.
A layered architecture from sensor to ERP
| Layer | Role | Example technology |
|---|---|---|
| Machine and controller | Data source | PLC, drives, energy meters |
| Edge | Read, preprocess, buffer | Industrial gateway, OPC UA, Modbus |
| Platform | Storage, rules, alerts | ThingsBoard, time-series database |
| Business systems | Planning and reporting | MES, ERP via API |
Two design decisions matter more than the choice of any single product. First, a consistent naming and data model, so that "machine running" means the same thing on a 1990s press and on a new CNC center. Second, a clear security boundary: the edge gateway reads from the production network and publishes outward, while no system in the office network writes directly to a controller.
I describe this structure in more depth in Smart factory architecture with IIoT.
Why operators decide whether it works
Machine data tells you that a machine stopped, not why. That is why I treat stop-reason categories, a simple input screen for operators and a display next to the line as part of the project, not as extras. When operators see their own data being used, data quality improves. For the broader organizational side of this change, see Digital transformation in traditional industries.
Frequently asked questions
Do PLC programs have to be changed to collect data?
Usually not. In most cases data is read only and control logic stays untouched. Any change that does become necessary is agreed with the plant's automation engineer and tested first.
What about machines without a PLC or network interface?
Those are instrumented with add-on sensors such as current transformers, vibration sensors or external counters. This approach is covered in detail under legacy equipment retrofit.
Is a smart factory the same as buying an MES?
Not necessarily. Many plants first need reliable machine data; an MES pays off once that foundation exists.
Can the project run without stopping production?
In most cases, yes. Installation is scheduled during planned maintenance windows or shift changes, and data collection runs in parallel with production.
