Engineering use augmented mixed virtual reality integrate artificial intelligence

What Is a Digital Twin?

Digital twins are virtual representations of real-world objects, systems, or environments, informed by real-world data inputs.

Digital Twin Key Takeaways

  • Digital twins help address a lack of data about assets or situations in business and industrial environments.

  • Digital twins use IoT sensors to create virtual representations of assets or environments with real-world data.

  • Digital twins help visualize current- and future-state environments in a contextualized way. With digital twins, businesses can optimize workflows, automate awareness, and predict future outcomes.

  • Businesses can use digital twins to contextualize and anticipate customer needs and provide quality engagements.

  • Intel hardware and software solutions help simplify digital twin deployments with powerful AI acceleration and software tools.



Digital Twins Address a Lack of Context in a Data-Driven World

The ubiquity of IoT cameras and sensors is generating a massive volume of data about assets and environments. Now the challenge is how to make sense of that data. Operators and administrators rely on manual methods to aggregate, understand, and act on scene data. However, this process is slow, and often the impacts of awareness and improvement aren’t immediately measurable. Decision-makers need tools to understand and contextualize a scene holistically, which will help them anticipate future problems and opportunities.
Digital twins offer a solution by capturing relevant data points about the work environment in an open, contextualized framework that allows people, systems, and processes to act on the data rapidly and efficiently. With a digital twin, businesses can stay one step ahead of potential problems and opportunities and help ensure positive, safe experiences for workers, customers, and assets.

How Does a Digital Twin Work?

A digital twin is a virtual abstraction of something in the real world, which can be represented either as collections of data points or as spatial recreations similar to what you see in a video game. The key difference between a digital twin and a simulation is that digital twins are created based on real-world data inputs and updated over time. Digital twins also serve as a foundation to measure the impact of multiple simulations, optimization efforts, and predictive models without needing to change anything in the physical world.

Benefits of Digital Twins

Digital twins give organizations deep visibility into assets, systems, and environments and provide the means to simulate changes or predict behaviors in a virtual realm. This experimentation opens the door to fast process improvement or adaptation on a vast scale. Digital twins can provide value to something as small as a local grocery store or something as large as a multifab semiconductor manufacturing plant.

As a solution, digital twins are still maturing and will support yet-unimagined use cases well into the future. However, digital twins are providing value right now in key areas:


  • Digital twins can help reduce the cost, time, and effort to change or optimize production workflows.
  • Digital twins can help keep assets and environments safe by establishing situational awareness to help monitor for anomalous behaviors and predict future outcomes.
  • Digital twins can help deliver positive engagements by enabling decision-makers to anticipate customer needs.
  • Digital twins can provide a standardized foundation of data to support other processes and services.
  • Real-time digital twins enable autonomy by providing a common fabric for coordinating and increasing the variety of ground- and air-based robots and vehicles.
  • Real-time digital twins enable better human interaction by enhancing what we see through augmented reality (AR) in the scene or through virtual reality (VR) when accessed remotely.

AI, Digital Twins, Awareness, and Anomalies

AI, machine learning, and deep learning can be used to apply a layer of cognitive decision-making to digital twin representations. Combining AI and digital twins helps automate situational awareness for a given asset or environment, especially when measuring conditions against historical patterns and trends to identify anomalous behavior.

For example, a digital twin of a grocery store floor plan could recognize long customer wait times at the checkout kiosk and notify staff to open more kiosks. The same digital twin could also potentially issue alerts when detecting excessive indoor movement after normal business hours.

Overcoming Data Siloes with Standardization

In the grocery store example, it’s important to note that the digital twin goes beyond a typical computer vision deployment in which the captured data may be siloed and only able to support specific or limited use cases. Because the digital twin is a virtual recreation of an asset or environment, it can be used to support a growing number of simultaneous applications. These applications can include people counting, social distancing, foot traffic pathing, inventory management, and product placement testing.

Digital twins offer a means to standardize data that is currently held in siloes by disparate systems and solutions. Rather than merely using AI computer vision to detect objects in a camera view, the digital twin opens contextualized scene data to any system or solution at the user’s behest.

Digital Twin Use Cases and Examples

Digital twins are more common than you think. Google Earth and Google Street View are both examples of digital twins—only with a rather slow update rate. Digital twins can track granular details, such as the temperature and vibration within a single appliance, or macro-level observations, such as the flow of vehicles through a busy intersection.

Wherever sensors are available to generate data about an asset or environment, a digital twin can theoretically aggregate the data into a virtual duplication of the real world. Digital twin operators can run as many parallel simulations as their edge or cloud servers allow to predict future outcomes and head off potential issues. Some key examples include:


  • Discrete manufacturing: Digital twins can help adapt product fabrication processes around downed equipment to minimize the impact on productivity. Digital twins can also make it possible to rapidly quote turnaround times for customer requests by simulating necessary changes to the production line.
  • Logistics: Digital twins can enable maritime port operators to rapidly visualize and simulate changes in container movements or vessel berthing schedules to optimize port efficiency.
  • Retail: Digital twins can drive automated situational awareness of environments while integrating with omnichannel platforms to drive product sales and customer engagement.
  • Healthcare: Patient monitoring tools can be aggregated in a digital twin to create a holistic portrait of patient health and map correlations in symptoms.

How to Create a Digital Twin

Digital twins typically require specialized software running on IoT edge server systems that pull real-world data from sensors and cameras. Alternatively, a solution can still be considered a digital twin if there is significant latency between data generation and ingestion, as in the case of Google Street View.

A digital twin does not need to be costly or overly complex. Many organizations can start with existing infrastructures and sensors simply by layering in some analytical tools. 
For example, a retail environment digital twin can still deliver value by tracing customer paths down to the meter, rather than the millimeter. Likewise, in logistics or manufacturing, a digital twin for situational awareness can start with basic proximity sensors installed on machinery.

Organizations can start with their existing infrastructure and add compute resources over time to improve the fidelity of the digital twin to the real world and improve outcomes over time. An incremental approach can lead to a virtuous cycle: more investment can generate more application value, which can generate more revenue, which can be invested back into the digital twin.

Intel® Hardware and Software Solutions

The latest generation of Intel® Xeon® Scalable processors for the IoT edge can help drive success for your digital twin deployments. Hardware-enabled AI acceleration and high core counts drive more parallelization for digital twin simulations and fast data movement between edge servers and external sensors. The Intel® portfolio also supports digital twin operators with key software offerings:


  • The Intel® Distribution of OpenVINO™ toolkit can help independent software vendors (ISVs) optimize their AI-driven digital twin solutions for fast performance on Intel® hardware.
  • Intel® Smart Edge is a software-defined edge computing platform that consolidates the orchestration of several sensors and tools across the IoT edge.

Digital Twins Help Bridge the Digital and Real Worlds

Digital twins help humans interact with the world better by offering new levels of transparency and visibility into assets and work environments that go beyond mere sensor data alone. With digital twins, solution providers can imagine entirely new ways to connect the physical world with machine learning models that can understand and act on real-world situations.

Frequently Asked Questions

A digital twin is a virtual representation of a real-world object or environment created and updated from real-world data. Digital twins draw on data from sensors and cameras to create a digital representation, which can then be used to drive awareness, anomaly detection, or optimization efforts.

Digital twins are still maturing and can be used across any industry. They have proven success in manufacturing by helping factory operators optimize their production lines, in logistics by helping ports run more efficiently, and in retail environments by helping store owners understand customer pathing.