Automation

Robots, PLCs, drives and sensors already produce the data needed to avoid downtime, reduce scrap and save energy. IoT makes it usable: machine data flows automatically into a platform instead of staying buried in the controller.

On this page you will find implemented IoT solution examples from automation, from condition monitoring of robot fleets and OEE measurement to connecting existing machines, each with end user, partner and result.

IoT in industrial automation: from the PLC to the cloud

Many automation systems are still islands: robots, drives and controllers produce reliably, but their data stays in the PLC. IoT connects machines, sensors and controllers to a data platform. Signals that already exist turn into metrics for maintenance, production and energy management.

Unlike many other industries, real-time behaviour, availability and functional safety are basic requirements in automation. The IoT layer must therefore not disturb the process: it reads along, analyses and only intervenes in the control system where that is explicitly intended, such as an automatic stop after a tool breakage.

Typical IoT applications in automation

Machine data acquisition and shop floor connectivity

Almost every project starts with machine data acquisition: status, cycle times, part counts and fault messages are read automatically from controllers instead of being written down by hand. Schaeffler built its own data acquisition platform that connects the machines of its 75 plants in a uniform way and paid for itself in less than a year. For machine tool builder CHIRON, machine connectivity is the foundation for remote services at customer sites.

Condition monitoring and predictive maintenance

The higher the degree of automation, the more expensive every stoppage. Sensors for vibration, temperature and current consumption reveal wear before a drive fails. WAGO uses this to determine the optimal maintenance time for the filters of a material distribution system, and ifm detects winding damage in fans purely from their current consumption. KUKA uses iiQoT to bundle the condition data of entire robot fleets. More examples in the use cases condition monitoring and predictive maintenance.

OEE monitoring and bottleneck analysis

Overall equipment effectiveness (OEE) is the product of availability, performance and quality. A line that runs 90 percent of planned time, reaches 95 percent of its target rate and delivers 98 percent good parts has an OEE of around 84 percent. With automatically captured data, this calculation becomes a live metric per machine and shift. At a Bosch Rexroth plant, ITK Engineering determines OEE with a portable transparency toolkit that connects machines from different vendors via OPC UA and MQTT.

Inline quality assurance

Quality defects that only show up at end-of-line testing are expensive. On honing machines, Schaeffler analyses vibrations in real time and stops the machine after a tool breakage before further parts are damaged. TURCK monitors the condition of solder paste in PCB production, and JUMO uses AI and IoT to increase yield. More in the use case quality management.

Commissioning, remote maintenance and service

For machine builders, IoT pays off after delivery too. Fabmatics organises commissioning through a service portal and cut the time required by 50 percent, while overall process speed rose by 30 percent. Pantec improves machine performance worldwide through remote access and data analysis.

Energy monitoring at machine level

Automated lines consume electricity, compressed air and cooling water. Measurements per machine show which system draws power while idle and where load peaks occur. This is the data basis for efficiency measures and robust CO₂ reporting. Doepke and Revolution Pi, for example, simplify the monitoring of residual currents. More examples in the use case energy monitoring.

OPC UA, MQTT and the asset administration shell: which standard for what

OPC UA is the standard for communicating with controllers: it delivers data with meaning, not just a value but also the fact that it is the spindle temperature in degrees Celsius. MQTT is a lightweight publish-subscribe protocol that distributes many data sources through a broker to many consumers, all the way to the cloud. The example from EMQ shows how such an MQTT broker scales. In practice the two are often combined: OPC UA at the machine, MQTT for transport.

The asset administration shell (AAS) goes one step further and describes a machine as a standardised digital twin with master data, documentation and operating data. In the Bosch Rexroth model factory in Ulm, it is used to reconfigure production lines flexibly, and XITASO connects existing machines through a function block.

Brownfield and retrofit: connecting existing machines

Most plants run new robots next to twenty-year-old controllers with Profibus, Modbus or proprietary interfaces. There are three ways to connect them: read the controller through a gateway, retrofit additional sensors that measure independently of the PLC, or combine both. Retrofitted sensors such as current transformers or vibration sensors are often the fastest route, because nobody has to touch the existing control logic. The example of RAFI and Device Insight shows how signal lamps and switches themselves become a data source.

IT/OT convergence and OT security

Connecting machines to IT systems and the cloud creates new attack surfaces. Industrial networks therefore need to be segmented, and external access should go through secured gateways. Rhebo and Bosch Rexroth show how anomaly detection can run directly on the controller ctrlX CORE, which is certified to IEC 62443. For wireless connectivity on the shop floor, Siemens relies on private 5G networks.

How to start an IoT project in automation

1. Start with the bottleneck: pick the machine or line whose downtime or scrap costs the most, not the newest one.

2. Define the metric: decide up front how success will be measured, for example OEE, unplanned downtime hours or scrap rate.

3. Clarify data sources: check what the controller already provides and where additional sensors are needed. Agree early with IT on how data may leave the OT network.

4. Pilot small: a pilot on one machine with a data acquisition kit delivers reliable figures within a few weeks and the basis for the business case.

5. Scale on standards: if you rely on OPC UA, MQTT and open platforms from the start, you can connect further machines and plants without reprogramming.

Finding IoT providers and partners for automation

The solution examples on this page come from providers who delivered their projects together with end users, including Device Insight, WAGO, ifm, TURCK, INSYS icom, Peakboard, Kontron AIS, Revolution Pi, Rhebo and EMQ. Each example shows the technology used, the starting point and the result. If you are more interested in connected products and services for machine builders, you will find further examples in the mechanical and plant engineering industry.

No marketing fluff. Only practice.