Looking at concrete digital twin examples is the fastest way to understand what a twin really is: a live virtual model kept in sync with a physical asset, process or place. In this article you will walk through twelve examples across manufacturing, energy, buildings, cities and healthcare, and for each one see the data sources, update rates and messaging patterns that make it work, so you can pick a realistic first project of your own.
If you want the full background first, read our digital twin technology guide. Here we stay practical.
What Every Digital Twin Example Has in Common
Before the list, it helps to know the four ingredients that appear in every working twin, regardless of industry:
- A physical counterpart that produces data: a machine, a building, a pipeline, a patient monitor.
- A data pipeline that moves that data reliably and with low latency. In most modern projects this is publish/subscribe messaging, and very often MQTT through an MQTT broker.
- A model that gives the data meaning: an asset hierarchy, a physics or statistical model, or simply a schema that says "this value is spindle speed in rpm".
- A consumer: a 3D view, a dashboard, an analytics job or a control loop that acts on the twin.
What differs between examples is the update rate, how much modelling sits on top of the raw data, and whether the twin only observes or also sends commands back. Keep those three dimensions in mind as you read.
Digital Twin Examples in Manufacturing
Manufacturing is where twins are most mature, because machines already expose rich data through PLCs and controllers. Our factory digital twin article covers the full plant view; here are four focused examples.
1. CNC Machine Twin
A CNC twin mirrors spindle speed, feed rate, axis positions, tool number, alarms and program state. Data typically comes from the controller through an edge gateway that publishes to topics such as acme/plant1/machining/line2/cnc4/telemetry. Update rates range from once per second for status to several times per second for axis positions if you animate the machine in 3D. The value lies in seeing every machine's real state in one place and correlating alarms with what the machine was doing at the time.
2. Robot Arm Twin
A robot arm twin streams joint angles, gripper state, payload and cycle counts. Because joint angles change quickly, this is one of the few examples where smooth animation demands higher rates, often 10 to 30 updates per second per robot. Engineers use it to replay faults, check reach and spot collisions in cell layouts before changing a program.
3. Industrial Furnace or Oven Twin
Thermal processes change slowly, so a furnace twin may only need zone temperatures, setpoints, burner state and door position every few seconds. The model layer is richer here: comparing actual curves against the recipe profile reveals drift long before parts go out of tolerance.
4. Conveyor and Production Line Twin
A line twin combines many small signals (motor running, jam sensors, part counters, photo-eyes) into a flow model of the whole line. It answers "where is the bottleneck right now?" rather than "what is one machine doing?". Line twins are a natural fit for an ISA-95 topic hierarchy where each cell and line has its own topic branch.
Digital Twin Examples in Energy and Utilities
5. Wind Turbine Twin
A wind turbine digital twin tracks rotor speed, pitch angle, yaw, generator temperature, vibration and power output. Turbines sit in remote places on constrained links, so efficient protocols and store-and-forward at the edge matter. The twin's main job is condition monitoring: comparing vibration and temperature against expected behaviour at the same wind speed.
6. Solar Plant and Battery Storage Twin
Inverters, string monitors, weather stations and battery management systems feed a plant-level twin. It combines measured output with irradiance to compute expected versus actual production, flagging soiled panels or failing strings. Battery twins add state of charge and cell temperatures so operators can plan charge and discharge windows.
7. Water Pumping Station Twin
Pump status, flow, pressure, tank levels and valve positions make up a pumping station twin. Unlike the previous examples, this one often closes the loop: operators change setpoints from the twin's interface and commands travel back to the site. That raises the bar for security and command acknowledgement.
Digital Twin Examples in Buildings and Cities
8. Smart Building Twin
A building twin maps HVAC units, room temperatures, CO2, occupancy and energy meters onto floor plans. Facility teams use it to find rooms that are heated while empty, or air handlers fighting each other. Data rates are low, but the number of points can be large, so a clear topic hierarchy per building, floor and room matters more than raw throughput.
9. Smart City Traffic and Infrastructure Twin
City twins combine traffic counters, signal timings, parking sensors, air quality monitors and public transport positions. A smart city digital twin is usually federated: each department owns its data, and the twin is a shared view built from many publishers. Pub/sub messaging suits this well because new consumers can join without changing the publishers.
10. Warehouse and Logistics Twin
Forklift positions, shelf occupancy, dock door status and order picking progress feed a warehouse twin. Its value is in spotting congestion and planning slotting changes. Location data arrives frequently, while inventory changes are event-driven.
Digital Twin Examples in Healthcare and Products
11. Hospital Equipment and Room Twin
Hospitals use twins of infusion pumps, imaging equipment, beds and room environments to track availability, utilisation and maintenance. Patient-facing data is regulated and needs strict access controls, so most early projects focus on equipment and environmental twins rather than clinical data.
12. Connected Product Twin
Manufacturers of compressors, chillers, vehicles or appliances keep a twin per shipped unit. The twin holds firmware version, configuration, usage hours and fault history, and supports remote diagnostics and over-the-air updates. This is the "device shadow" style of twin; see device shadow vs digital twin for how the two concepts differ.
Comparing the Digital Twin Examples
| Example | Typical data sources | Update rate | Observes or controls |
|---|---|---|---|
| CNC machine | Controller via edge gateway | 1–10 per second | Mostly observes |
| Robot arm | Robot controller joint data | 10–30 per second | Observes, some commands |
| Furnace | Thermocouples, burner PLC | Every few seconds | Observes, recipe checks |
| Production line | Sensors, counters, PLCs | 1 per second, events | Observes |
| Wind turbine | SCADA, vibration sensors | 1 per second or slower | Observes |
| Solar and storage | Inverters, BMS, weather | Every few seconds | Observes, dispatch plans |
| Pumping station | RTUs, flow and level sensors | Seconds | Controls setpoints |
| Smart building | BMS, IoT room sensors | Minutes | Observes, some control |
| Smart city | Counters, air quality, GPS | Seconds to minutes | Mostly observes |
| Warehouse | RTLS, scanners, door sensors | Sub-second to events | Observes |
| Hospital equipment | Device gateways, BLE tags | Seconds to minutes | Observes |
| Connected product | Embedded device firmware | Events and periodic | Observes and configures |
The Data Pattern Behind Most Digital Twin Examples
Look closely and almost every example follows the same messaging pattern. Devices or gateways publish telemetry to a structured topic, the latest known state is kept available for anyone who connects late, and consumers subscribe with wildcards to the branch they care about. MQTT gives you all three natively: hierarchical topics, retained messages and wildcard subscriptions. That is why MQTT has become the default transport for twins; we go deeper in digital twin with MQTT.
You can try the pattern in a minute against the public MQTT test broker. Pick a unique prefix and publish a retained state message for an imaginary pump:
mosquitto_pub -h broker.connectmq.com -p 1883 -i twin-demo-pub-7f3a \
-t "connectmq/demo/7f3a/site1/pumps/p1/state" -r \
-m '{"running":true,"flow_lpm":412,"pressure_bar":3.8}'
Then subscribe from another terminal. Because the message is retained, you get the current state immediately, exactly as a twin viewer would on page load:
mosquitto_sub -h broker.connectmq.com -p 1883 -i twin-demo-sub-7f3a \
-t "connectmq/demo/7f3a/site1/#" -v
The public broker is readable by anyone, so use it only for test data. For real assets, a managed MQTT broker with credentials and TLS is the right foundation.
Choosing Your First Digital Twin Project
Not every example is a good starting point. Use this checklist to pick a first twin that delivers value quickly:
- Data already exists. Prefer assets that already expose data through a PLC, gateway or API.
- One clear question. "Why does line 2 stop more often than line 1?" beats "digitise the plant".
- Observe before you control. Start read-only; add commands once data quality and security are proven.
- Moderate update rates. One update per second per asset is easy to handle end to end.
- A visible consumer. A dashboard or 3D view that operators actually open builds support for the next phase.
ConnectMQ runs a 3D digital twin web demo of a small factory with a CNC machine, robot arm, furnace, tank, conveyor and pump, all updated live over secure WebSocket. It is a useful reference for how examples 1 to 4 look when wired together.
Build Your Own Digital Twin on ConnectMQ
Every example above starts with reliable real-time messaging. ConnectMQ gives you a cloud MQTT broker with MQTT 3.1.1 and 5.0, TLS on 8883 and secure WebSocket on 8084 for browser-based 3D views, plus a live topic explorer to watch your twin's data arrive. Create a broker on the free Serverless plan at app.connectmq.com, with no credit card required, and connect your first asset today.
Frequently Asked Questions
What is the simplest digital twin example to start with?
A single machine or room with a handful of sensors published over MQTT and shown on a dashboard. It proves the data pipeline and gives you a template to copy to more assets.
Do digital twin examples always need 3D models?
No. Many valuable twins are dashboards, tables or state documents. 3D helps with spatial problems like layouts and robot reach, but the data model and live state matter more.
Which industries use digital twins the most?
Manufacturing, energy and utilities, buildings, logistics and cities are the most common, with healthcare growing mainly around equipment and facilities.
Why is MQTT used in so many digital twin examples?
MQTT offers lightweight publish/subscribe, hierarchical topics and retained messages, which map directly to streaming state, organising assets and giving new viewers the latest values instantly.
