Using Digital Twins
This overview of digital twins will help explain where they fit into an IIoT ecosystem and how the data they furnish is critical to success.
Join the DZone community and get the full member experience.Join For Free
Recently, I was invited by Bruce Sinclair from IoT-inc.com to discuss digital twin technology in his podcast series. We went through a hour-long discussion and covered many aspects of this emerging, game-changing technology. We explored many dimensions of the digital twin, including its origins, how it is built and used, the technologies behind it, the business value it now delivers and its future potential, and some of the outcomes digital twins can provide for industrial organizations.
In this blog post, I want to recap three major aspects from the podcast:
- Define digital twin
- Showcase the outcomes you should expect from digital twins
- Discuss the technologies behind digital twins
What Is a Digital Twin?
A digital twin is a software construct that bridges physical systems and the digital world. Think about it as software being paired to an instrumented physical object (or system or machine). The digital twin will be able to act as a proxy to the digital world. It accumulates data over time about the structure of the system, its operation, and the environment in which it operates. Together with the data, intelligence is built on top using analytics, physics, and machine learning. You can query the digital twin of a specific system and ask about past and present performance and operations, and also ask for early warnings and predictions.
What Outcomes Should You Expect From Digital Twins?
Let’s be clear, digital twins are not an academic modeling exercise! They are built for specific outcomes. We first think about the desired outcome — a specific key performance indicator (KPI) to ensure a specific quality of service, or a prediction of the life of a specific part in order to minimize downtime, etc. — and then we figure out the bundle of data and intelligence we need to deliver that outcome.
The progression of value usually follows this path:
Understand operation and prediction at the asset level and leverage to optimize individual performance.
Optimize maintenance at individual level.
Aggregate for multiple assets and optimize at the operations level.
Rethink business models and offer new value and services.
What Are the Key Technologies?
Digital twins start with data and metadata: the asset model, the sensors, and actuator data, as well as very broad context data or knowledge that is related to the design, building, operation, and servicing of the physical twin. Analytics are the brain of the digital twin. Intelligence is built on top of the data using analytics, physics, and algorithms or machine learning. The digital twins live on a platform, which ensures persistence of the digital twin, pairing between the digital and physical twins, and creates a learning system to continuously improve the fidelity of the digital twins it hosts. Finally, the apps deliver the ultimate outcome by consuming the value generated by the digital twin, usually using APIs to query the digital twin.
Published at DZone with permission of Dimitri Volkmann, DZone MVB. See the original article here.
Opinions expressed by DZone contributors are their own.