# Industry 4.0

Industry 4.0 describes the connected factory, where machines, sensors, software and AI share data to improve quality, efficiency and flexibility. Learn its building blocks.

## Overview

Industry 4.0 is the name for the fourth industrial revolution: factories and supply chains in which machines, sensors, control systems, enterprise software and analytics are connected and share data. The idea is that a production system should be able to sense its own state, analyse it and adapt, rather than run as a set of isolated machines managed by paper, spreadsheets and experience.

In practical terms, Industry 4.0 covers connected equipment, the industrial internet of things (IIoT), data platforms that join operational and business data, and applications such as [predictive maintenance](/tools/predictive-maintenance), automated quality inspection, [digital twins](/tools/digital-twins) and flexible scheduling. Its goals are higher quality, less downtime, lower energy and material use, faster changeovers and products tailored to individual orders.

The name comes from a simple history. The first industrial revolution brought steam and mechanisation, the second electricity and mass production, and the third electronics and computer-controlled automation. The fourth connects these automated systems into networks that exchange data and make decisions, much as a city is transformed not by building more roads but by linking traffic lights, buses and maps into one responsive system.

## What it is

| Building block | Role |
|---|---|
| Sensors and actuators | Measure physical conditions and act on equipment |
| PLCs and control systems | Run machines in real time |
| SCADA | Supervises and visualises processes across a site |
| MES | Manages production orders, tracking, quality and performance on the shop floor |
| ERP | Plans resources, orders, finance and supply across the business |
| Industrial communication | Protocols such as OPC UA and MQTT that move data between layers |
| Edge computing | Processes data close to machines for speed and bandwidth |
| Data platform | Stores and integrates time series, events and business data |
| Analytics and AI | Detects, predicts and optimises from the data |
| Cyber-physical system | A physical process tightly integrated with computation and networking |

A useful mental model is a layered stack. At the bottom are field devices; above them control; then supervision; then manufacturing operations; and at the top enterprise planning. The ISA-95 standard describes these layers and how information should flow between them. Industry 4.0 aims to break down the walls between layers so that data from a sensor can inform a business decision, and a change in a customer order can reach a machine quickly.

## Where it came from

The term "Industrie 4.0" originated in Germany. According to one of its originators, Wolfgang Wahlster of the German Research Center for Artificial Intelligence (DFKI), he proposed the name at a meeting in December 2010 with Henning Kagermann and Wolf-Dieter Lukas, as a more intuitive alternative to "cyber-physical production systems". The three co-authored an article introducing the concept in April 2011, and the term was presented publicly at the opening of the Hannover Messe trade fair that same month.

The idea became part of German industrial and research policy, and a national Plattform Industrie 4.0 was formed to coordinate industry, unions, research and government. In 2015 the platform, with industry associations, presented the Reference Architectural Model Industrie 4.0 (RAMI 4.0), a framework for describing Industry 4.0 systems, which recommended OPC UA for the communication layer.

Similar programmes appeared elsewhere under other names, such as smart manufacturing and the industrial internet. In recent years, discussion has also turned to "Industry 5.0", which places more emphasis on human-centred work, resilience and sustainability alongside automation.

## How it works

### Step 1: connect the equipment

Machines are fitted with sensors or have their existing controller data exposed. Older equipment, often decades old, may need retrofit sensors and gateways. Data are exposed through standard protocols, increasingly OPC UA, which also carries information models describing what each value means.

### Step 2: collect and contextualise

Edge gateways gather data, filter it and attach context: which machine, which order, which product. Data flow into a historian, a time-series database or an event platform such as [Apache Kafka](/tools/apache-kafka). A unified naming scheme, sometimes called a unified namespace, keeps data findable across the plant.

### Step 3: integrate operations and business data

Shop-floor data are joined with MES and ERP data so that sensor readings can be linked to orders, batches, materials and costs. This is where many projects stall, because operational technology (OT) and information technology (IT) have different owners, tools and priorities.

### Step 4: analyse and model

On the integrated data, teams build applications: overall equipment effectiveness (OEE) dashboards, [anomaly detection](/tools/anomaly-detection), computer vision for defects, demand and throughput forecasts, and optimisation of schedules and settings. Machine learning operations practices from [MLOps](/topics/mlops-ai-infrastructure) keep models reliable in production.

### Step 5: act

Results go back to people as alerts, work orders and recommendations, or to systems as updated schedules and set points. Changes that affect equipment usually require human approval and strict safety checks.

### Step 6: secure and govern

Connecting machines exposes them to cyber threats. Network segmentation, strong identity, encrypted connections and monitoring are essential. Standards such as the IEC 62443 series address industrial cybersecurity.

## Key concepts and terms

| Term | Meaning |
|---|---|
| IIoT | The industrial internet of things: networked sensors and machines |
| Cyber-physical system | Physical processes controlled and monitored by integrated computing |
| OT and IT | Operational technology that runs equipment, and information technology that runs business systems |
| OPC UA | A platform-independent standard for industrial data exchange and information modelling |
| MQTT | A lightweight publish-subscribe messaging protocol common in IoT |
| MES | Manufacturing execution system for managing production on the shop floor |
| OEE | Overall equipment effectiveness: availability × performance × quality |
| Edge computing | Processing data near its source |
| Unified namespace | A shared, structured place where all plant data are published |
| RAMI 4.0 | A German reference architecture model for Industry 4.0 |
| Retrofit | Adding sensors or connectivity to existing equipment |

## A worked example

This scenario is illustrative. A mid-sized plastics moulding company runs 30 injection-moulding machines of various ages. Scrap is about 4 per cent, and planners learn about machine stoppages only when an operator phones.

In phase one, the company adds edge gateways that read cycle data, temperatures and pressures from each machine controller, plus simple retrofit sensors on the oldest machines. Data flow into a central time-series store tagged by machine, mould and order. Within weeks an OEE dashboard shows that availability averages 71 per cent, with mould changes the largest loss.

In phase two, the team links process data to quality results from the MES. A [gradient boosting](/tools/gradient-boosting) model finds that scrap rises sharply when mould temperature drifts more than a few degrees from its set point early in a shift. Operators receive alerts when drift begins. Over six months scrap falls to 2.8 per cent, and standardised changeover procedures, prompted by the dashboard, lift availability to 77 per cent. The company then plans a third phase on scheduling.

## Strengths and limitations

**Strengths:**

- Makes production visible in real time across machines and sites.
- Links shop-floor data to orders, costs and customers.
- Enables predictive, rather than reactive, operations.
- Supports flexible production and smaller batch sizes.
- Creates a data foundation for continuous improvement.

**Limitations:**

- Integration of old equipment and many protocols is slow and costly.
- OT and IT organisations must work together, which is often hard.
- Connected factories face real cybersecurity risks.
- Value comes from changed processes, not technology alone.
- Smaller manufacturers may lack skills and budgets.

## How it compares with alternatives

| Approach | Scope | Data use | Typical outcome |
|---|---|---|---|
| Industry 4.0 | Whole value chain, connected | Integrated, real time | Adaptive, data-driven operations |
| Traditional automation (Industry 3.0) | Individual machines and lines | Local, siloed | Efficient but rigid production |
| Lean manufacturing | Processes and people | Manual metrics | Less waste, often a foundation for 4.0 |
| [Smart manufacturing](/tools/smart-manufacturing) | Shop-floor applications | Integrated | Largely overlapping term, used widely in the US |
| Industry 5.0 | Human-centred, sustainable industry | Builds on 4.0 | Emphasis on people, resilience, environment |

## Practical guidance and best practices

1. Start from business problems, such as scrap or downtime, not from technology.
2. Pilot on one line, measure the result, then scale.
3. Agree a data model and naming scheme early, before connecting hundreds of devices.
4. Prefer open standards such as OPC UA and MQTT to avoid lock-in.
5. Form joint OT and IT teams with shared goals.
6. Design security into every connection from the first pilot.
7. Train operators and involve them in designing alerts and dashboards.
8. Track a small set of metrics, such as OEE and scrap, before and after each change.

## Common mistakes

- Buying a platform before defining the problems it should solve.
- Collecting everything at high frequency without a plan to use it.
- Connecting machines to corporate networks without segmentation.
- Running endless pilots that never scale.
- Ignoring the people who must change how they work.

## Frequently asked questions

### Is Industry 4.0 the same as automation?

No. Automation makes individual machines or lines run without manual effort, which was the main theme of the third industrial revolution. Industry 4.0 connects automated systems with each other and with business systems so they can share data and adapt.

### Do I need to replace old machines?

Usually not. Many old machines can be connected through retrofit sensors and gateways. Replacement may make sense for other reasons, but connecting existing equipment is often the fastest way to start.

### Where does AI fit into Industry 4.0?

AI turns the connected data into decisions: detecting defects in images, predicting failures, forecasting demand, optimising settings and schedules, and increasingly answering operators' questions in natural language. It depends on the data foundation the rest of Industry 4.0 provides.

### What is OPC UA and why does it matter?

OPC UA is an open standard for exchanging industrial data securely between devices and software from different vendors, with built-in information models that describe what the data mean. It was recommended for the communication layer in RAMI 4.0 and is widely used in connected factories.

### What is Industry 5.0?

It is a newer framing that keeps the technologies of Industry 4.0 but puts more weight on human-centred work, resilience to disruption and sustainability. It is better seen as an evolution of priorities than as a separate technical revolution.

## Glossary

- Industry 4.0: the connected, data-driven fourth industrial revolution.
- PLC: a programmable logic controller that runs machine control logic.
- SCADA: supervisory control and data acquisition for monitoring processes.
- ERP: enterprise resource planning software.
- Historian: a database for industrial time-series data.
- OEE: a standard metric combining availability, performance and quality.
- Gateway: a device that connects field equipment to networks.
- Segmentation: dividing networks into zones to limit the spread of attacks.

## Summary

Industry 4.0 describes factories and supply chains in which machines, control systems, business software and analytics are connected and share data so production can sense, analyse and adapt. The term was coined in Germany in 2010 and presented publicly in 2011, and frameworks such as RAMI 4.0 followed. Its building blocks are sensors, industrial protocols, edge computing, integrated data platforms and AI applications. Success depends on solving concrete problems, open standards, joint OT and IT work, strong security and bringing people along.