IoT in industrial automation: how it works and where it pays off
Head of Innovation
Digital Transformation, AI, IoT, Machine Learning, and Cloud Technologies. Nearly 15 years driving innovation

IoT in industrial automation means connecting sensors, PLCs and machines to a network so their data can be collected, analysed and acted on, often automatically. Classic automation follows fixed logic on the factory floor. IoT adds a data layer on top, which enables condition monitoring, predictive maintenance, energy tracking and remote control of whole sites.
Key takeaways
- IoT does not replace PLCs and SCADA. It reads data from them and from added sensors, then sends it to edge or cloud systems for analysis.
- The most proven use cases are condition monitoring, predictive maintenance, energy monitoring, quality inspection and asset tracking.
- OPC UA and MQTT are the common protocols for getting machine data out in a standard form.
- Security is the main risk. NIST SP 800-82 and IEC 62443 are the reference frameworks for operational technology (OT).
- EU rules now matter: the Data Act applies from 12 September 2025 and the Cyber Resilience Act's main obligations from 11 December 2027.
- Start with one line or one asset class and a measurable goal, not a plant-wide platform.
How does IoT enable automation in industry?
It closes the loop between what machines are doing and the decisions made about them. Sensors measure, networks move the data, software finds patterns, and either people or control systems act on the result.
Traditional automation already runs machines well. A PLC opens a valve when pressure hits a threshold, and a SCADA system shows the operator what is happening. What it usually lacks is history, context and reach. Data stays on the line, gets overwritten, and nobody compares one plant with another.
Industrial IoT changes that by adding four layers:
- Devices and sensors: vibration, temperature, current, pressure, flow, cameras. Some are new; many signals already exist in PLCs.
- Connectivity: wired Ethernet, Wi-Fi, private 5G or low-power radio, plus a gateway that translates industrial protocols.
- Edge and cloud processing: filtering and alarms near the machine, storage and analytics in the cloud.
- Applications and actions: dashboards, work orders in the maintenance system, set-point changes, or alerts to the right person.
Our explainer on IoT integration goes deeper into how these layers connect to ERP and MES systems.
The architecture in practice
Most industrial IoT systems follow the same pattern, whatever the vendor.
Getting data off the machines
Machine data is first read from controllers and sensors. OPC UA, maintained by the OPC Foundation, is the main vendor-neutral standard for exchanging industrial data with a defined information model. MQTT is a lightweight publish/subscribe protocol that suits unreliable networks and many small messages. A typical design uses OPC UA on the plant floor and MQTT from the gateway to the cloud.
Edge versus cloud
Not every decision should wait for the cloud. Safety interlocks and fast control loops stay in the PLC. Filtering, local alarms and buffering during network outages run on an edge device. Long-term storage, cross-site comparison and model training run in the cloud.
On AWS, for example, AWS IoT Greengrass runs code on edge devices, and AWS IoT SiteWise collects and models equipment data at scale. Azure and other platforms offer similar building blocks.
| Task | Where it should run | Why |
|---|---|---|
| Safety stops, fast control loops | PLC / controller | Millisecond timing; must work without a network |
| Filtering, local alarms, buffering | Edge gateway | Cuts data volume; keeps working during outages |
| Dashboards across sites, history | Cloud | Central view, cheap long-term storage |
| Training predictive models | Cloud | Needs months of data and compute on demand |
| Running trained models on the line | Edge (sometimes cloud) | Low latency for inspection or anomaly alerts |
Where IoT in automation pays off
Pick use cases where a measurable cost already exists: unplanned downtime, scrap, energy or manual rounds. These five are the most established.
Condition monitoring and predictive maintenance
Vibration, temperature and motor current reveal wear on bearings, pumps, fans and gearboxes before they fail. The first step is condition monitoring with simple thresholds. Predictive maintenance adds models that estimate time to failure, which needs failure history to train on. See our guide to the benefits of predictive maintenance in industry for how to build the business case.
Energy monitoring
Sub-metering per line or machine shows where energy goes and when. It often finds idle machines running overnight, compressed-air leaks and peaks that drive grid tariffs. With energy reporting duties growing in the EU, the same data also feeds sustainability reports.
Quality inspection
Cameras and computer vision models check parts on the line and flag defects in real time. The result can stop a line, divert a part or adjust a machine setting. This is one of the clearest examples of IoT data driving an automatic action. Opsio's visual inspection service covers this use case.
Asset and material tracking
RFID, BLE tags and GPS show where tools, pallets, containers and vehicles are. It cuts search time and lost assets, and gives planners a live picture of work in progress.
Remote monitoring of distributed assets
Pumping stations, wind turbines, chargers and machines installed at customer sites can be monitored and sometimes controlled remotely. For equipment makers this also enables service contracts based on actual usage.
What are the challenges?
The technology is mature; the hard parts are security, integration and organisation. Plan for these from day one.
OT security
Connecting machines that were designed for isolated networks creates new attack paths. Many PLCs have no authentication and cannot be patched quickly. NIST SP 800-82 Rev. 3 is the standard guide to OT security, and the ISA/IEC 62443 series defines zones, conduits and security levels for industrial systems.
Practical minimums:
- Separate OT and IT networks, with a controlled zone (DMZ) between them.
- Outbound-only connections from gateways to the cloud; no inbound ports to the plant.
- Device certificates, not shared passwords, for every gateway and sensor that talks to the cloud.
- An asset inventory of every connected device, including firmware versions.
- Monitoring of OT traffic for unexpected connections.
Interoperability
A plant may run controllers from several vendors and generations. Some speak OPC UA, others Modbus, Profinet or a proprietary protocol. Budget for gateways and protocol conversion, and agree a common data model (asset names, units, tags) before you connect the first machine.
Data quality and ownership
Sensors drift, timestamps disagree and tags get renamed. Without clear ownership, dashboards lose trust within months. Name a data owner per site and add basic checks for missing or out-of-range values.
Regulation
Two EU laws affect industrial IoT directly. The EU Data Act has applied since 12 September 2025 and gives users of connected products rights to access the data those products generate. That matters when a machine vendor's cloud holds your production data. The Cyber Resilience Act sets security requirements for products with digital elements; its reporting obligations apply from 11 September 2026 and its main obligations from 11 December 2027.
How to start an industrial IoT project
Start small, with one problem that has a clear cost. A focused pilot proves value and teaches you the integration work before you scale.
- Choose one use case and one asset class. For example, vibration monitoring on ten critical pumps.
- Define the success measure. Fewer unplanned stops, hours of manual rounds saved, or kWh per unit produced.
- Map existing data. Many signals already exist in PLCs and historians; add sensors only where gaps remain.
- Design security first. Network zones, gateway identity and data flows, agreed with whoever owns OT.
- Build edge-to-cloud once. A reusable pipeline makes the next use case far cheaper.
- Connect to work processes. An alert that does not create a work order or reach a named person will be ignored.
- Review after three to six months. Decide to scale, change or stop based on the measure from step 2.
How Opsio helps
Opsio designs and runs industrial IoT platforms on AWS and Azure, from edge gateways to cloud data pipelines and dashboards. Our IoT solution development team handles architecture, security and integration, and our IoT predictive maintenance service covers condition monitoring through to failure prediction.
Frequently asked questions
What is the difference between IoT and industrial automation?
Industrial automation uses controllers such as PLCs to run machines by fixed logic. IoT connects those machines and extra sensors to networks and software, so data can be stored, analysed and used across sites. Automation acts locally; IoT adds visibility and data-driven decisions on top.
What is IIoT?
IIoT, the Industrial Internet of Things, is IoT applied to manufacturing, energy, utilities and logistics. It differs from consumer IoT in its focus on reliability, long equipment lifetimes, industrial protocols and safety.
Which protocols are used in industrial IoT?
OPC UA is the main standard for structured machine data. MQTT is widely used to send data from gateways to the cloud. Plants also run Modbus, Profinet, EtherNet/IP and vendor protocols, which gateways translate.
Is industrial IoT secure?
It can be, if designed with OT security in mind. Use network segmentation, outbound-only connections, device certificates and asset inventories, following NIST SP 800-82 and IEC 62443. Most incidents come from flat networks and remote-access tools left open.
How much data does an industrial IoT system produce?
It depends on signal count and sample rate. A temperature reading once a minute is tiny; high-frequency vibration data from one sensor can reach gigabytes per day. Edge filtering and sending features rather than raw waveforms keep cloud costs under control.
Related Articles
About the Author

Head of Innovation
Jacob leads innovation at Opsio, specialising in digital transformation, AI, IoT, and cloud-driven solutions that turn complex technology into measurable business value. With nearly 15 years of experience, he works closely with customers to design scalable AI and IoT solutions, streamline delivery processes, and create technology strategies that drive sustainable growth and long-term business impact.
Editorial standards: This article was written by a certified practitioner, named in the byline above. We revise it when the technology or regulation it covers changes; the dates above show the latest update. Opsio maintains editorial independence — we recommend solutions on technical merit, not commercial relationships.