Digital Twins: Your Next Step in IoT

Digital twins bridge the gap between the physical and virtual worlds and help you realise the full potential of your Internet of Things (IoT) solutions. Now more accessible than ever, digital twins are the next step in your IoT journey.

What are Digital Twins?

Digital twins are a digital representation or model of a physical thing enabled by IoT.

IoT is the network of physical devices (things) connected to the internet, collecting data and telemetry. An IoT device could be your fridge, your Alexa, a truck, a piece of machinery in a factory, or a building. You can integrate almost anything with IoT technology.

Digital twins are a map of IoT enabled devices showing how they connect and work together, supplemented by other business information and collected data. A digital twin acts as a blueprint highlighting the relationships between each piece and allowing for complex data analysis and action.

A digital twin of an object should have:

  1. A model of the object
  2. An evolving set of data relating to the object, and
  3. A means of dynamically updating or adjusting the model in accordance with the data.

(Source 1)

Say, for example, you have an office building with IoT enabled thermostats in each room. The thermostats and their data alone would not be digital twins. However, overlay that information into a building model and create a means of updating the temperature change to reflect the real world, and you have digital twins. The model could be a graph, a dashboard, a 2D map, or even a 3D virtual replica of the building.

Tweet from Frame VR showing a digital twin of the Trello office

Why Digital Twins?

A majority of all IoT projects are considered failures. According to a report by Beecham Research titled “Why IoT Projects Fail”, only 12% are considered entirely successful. The projects fail because businesses focus on single-point solutions, jumping on new technology trends, and not looking at the big picture. As a result, projects end up with islands of things instead of an internet of things. The data is not linked together, leaving each IoT solution isolated on an island.

Connecting everything creates a complete picture that can change how we look at the problem. We can make these connections with digital twins.

Digital twins turn our islands of things back into an internet of things through relationships and models. We create insights from the mass amounts of data collected, revealing practical solutions and allowing for action with digital twins.

Actions from digital twins include:

  • Analysis and visualisation of the current state
  • Design and validation of processes new or existing
  • Simulation of scenarios
  • Increased safety and reliability with alerting
  • Optimisation of existing processes
  • Tracking and reviewing historical information
  • Predictive performance and maintenance
  • Real-time control over the physical twin

(Source 2)

Taking data-driven action in our IoT solution can turn a failure into a success.

History of Digital Twins

The concept of digital twins is not new. The term ‘Digital Twins’ was first used in 2002 by Michael Grieves of Florida Institute of Technology. (Source 3)

In theory, digital twins date back even farther. The first use of digital twins concepts dates back to the ’60s with NASA. There is a theory that Apollo 13 was the first example of digital twins. NASA had complex simulation environments to model the spacecraft powered by computer systems, technology, and algorithms. Mission control used these simulators and the telemetry coming in from the ship to quickly adapt and model the damaged spacecraft, allowing them to get the crew home safely. (Source 4)

NASA was not using IoT devices at the time, but the theories of using telemetry to update models and make decisions are what we use now with digital twins.

Apollo 13 Mission Control with Fred Haise and Gene KranzApollo 13 Mission Control with Fred Haise and Gene Kranz

Digital Twins Now

The difference with digital twins now is that we can get information in real-time. Real-time data allows us to react and automate in real-time. Then, using the power of the cloud, we can feed the stream of data into real-time processing, alerting, machine learning algorithms, advanced insights, and automation.

The cloud has also made digital twin technology more accessible and easy to use. No longer do you need a supercomputer and technology of NASA to create advanced simulations and digital twins. Instead, cloud providers take care of the compute for us, allowing us to focus on innovative solutions.

With broader accessibility, we can use digital twins anywhere you have IoT devices. Typical categories include consumer, commercial, industrial, and infrastructure. Examples usually include smart buildings and industrial use cases, but the opportunities are endless.

Digital twins power the freeways that use digital speed limits and change based on traffic patterns. Health monitors can leverage digital twins to send alerts to the user or a doctor. A bakery can leverage digital twins to track deliveries, add predictive maintenance to the ovens, and use video analytics to track food waste. You can even create a digital twin of your own home.

Examples of industries that can use digital twins: farming, city, bakery, medical, racecar, industrialExamples of industries that can use digital twins: farming, city, bakery, medical, racecar, industrial

How to get started

As a software engineer in the Microsoft space, I recommend starting with Azure IoT. Azure IoT provides a broad range of services and capabilities suitable across industries and includes the Azure Digital Twins platform.

Azure Digital Twins integrates seamlessly with the other Azure services and enables the creation of knowledge-based graphs based on digital models of entire environments.

The core capabilities provided in the Azure Digital Twins platform are:

  • Open Modeling Language - Digital Twin Definition Language (DTDL)
  • Live Execution Environment - Azure Digital Twins Explorer
  • Input from IoT and Business Systems
  • Output to Time Series Insights, Storage, and Analytics

In addition to the robust capabilities, Microsoft provides a detailed suite of documentation and learning modules to help you get up and running.

Check out digital twins and take your next step in IoT.


References

  1. Wright, L., & Davidson, S. (2020). How to tell the difference between a model and a digital twin. Advanced Modeling and Simulation in Engineering Sciences, 7(1), 1–13. https://doi.org/10.1186/s40323-020-00147-4

  2. Singh, M., Fuenmayor, E., Hinchy, E. P., Qiao, Y., Murray, N., & Devine, D. (2021). Digital Twin: Origin to Future. Applied System Innovation, 4(2), 36. https://doi.org/10.3390/asi4020036

  3. Grieves, M. W. (2019). Virtually intelligent product systems: digital and physical twins. https://doi.org/10.2514/5.9781624105654.0175.0200

  4. Ferguson, S. (2020). Apollo 13: The First Digital Twin. Siemens® https://blogs.sw.siemens.com/simcenter/apollo-13-the-first-digital-twin.


Melissa Houghton

Melissa Houghton is a Lead Software Engineer at Azenix and a Microsoft MVP in Developer Technologies.

Her work is focused on application development using Angular, .NET, and Azure, but she has a wide variety of skills and is always open to learning new things.

An advocate for women in tech with a passion for leadership, technology and giving back to the community. She frequents tech events, is an international conference speaker and organiser of DDD Melbourne conference, Melb․NET meetup, and Microsoft AU New Breakpoint Community.

Originally from California, Melissa loves to travel and now lives in Melbourne, Australia.