Blog
With every new technology comes hype and buzzwords. In the last decade, we’ve seen hype around everything from the Internet and business intelligence to mainframes and mini-computers. These days, the spotlight is on the Internet of Things (IoT), and industry is buzzing around digital twin specifically.
In 2016, I first started tracking mentions of digital twins online. I setup a Google Alert for the term and initially received two to four alerts monthly. Fast forward to a year later, and I pretty much get alerts daily. In fact, Gartner has listed digital twins in its Top 10 Strategic Technology Trends for 2017.
The earliest reference to digital twin that I could find is from 2002, and it is anchored in the Product Lifecycle Management space, specifically for manufacturing complex systems such as spacecraft. The initial concept was to create a digital model of a physical system before building it. This way, tests and simulations could be performed beforehand, thus validating design upfront and ultimately saving effort, time, and money in the long run. This phase of the digital twin concept—when it only exists in the digital world—is often referred to as “Digital Twin Prototype”, “Digital Twin Class,” or “Digital Twin Blueprint."
Moving from designing models or blueprints to actually building the object, you start to transition from the digital to the physical world. This is where the first system is built. This system, although based on a specific design, will have its own characteristics. For instance, parts are built according to tolerances, but can vary since two physical objects never have the exact same size or dimensions. You can also digitize it, capturing the data and adding it to the model. Now we are talking about a “Digital Twin Instance,” which has the specifics of an individual object—the digital representation is now paired to an actual physical object or asset.
The physical world has already gone through a revolution—especially in industrial use cases—where we can instrument machines and equipment with sensors and actuators. Industrial organizations can now monitor and control systems digitally. By connecting this to digital twins, organizations can also start to accumulate information about the operation of the physical system it is connected to. Then take it a step further—add more data to enrich the digital twin by capturing environmental data, such as location and configuration, and more general information, like services record, financial models, etc.
The result: industrial organizations end up with a digital construct that knows everything about the design, the building or manufacturing, the past and current operations, and the servicing of the physical system. If intelligence is added in the form of analytics, models, and other algorithmic techniques, such as machine learning, organizations can start receiving predictions and early warning alerts faster than ever before.
A digital twin acts as a proxy to immerse physical systems into the digital space, giving access to structure, context, and behavior of an asset. It provides you with a window to past and present states and conditions, and gives you the ability to look into the future.
Digital twins don’t just need to act independently, sometimes a group of digital twins, or an “aggregate”, can also be beneficial. If your organization is monitoring multiple systems of the same type of assets, for instance a fleet of jet engines (each of which has an individual digital twin), you can start to learn from all of them as a cohort, find similar patterns or trends, and that analysis can lead to refining models for higher fidelity in the future.
By bringing all digital twins onto a common platform, you can effectively build a learning system for the physical world. From a simple concept of building a digital model of an object before an asset is built, to analyzing the data from a cohort of digital twins on a fleet of assets, a powerful digital construct has developed that can lead to significant insights for your asset-centric organization
Common use cases for digital twin technology today include, machine and equipment heath (are my systems working fine today?) as well as predictive maintenance (how do I eliminate unplanned downtime of my machines?). But this is just the beginning, now that you better understand your physical systems digitally, you can start combining and orchestrating all the digital twins of your operations to improve global performance and likely create new value and new business models. From design to build and operate to servicing and decommissioning—a digitalized product lifecycle can be enabled across your entire value chain, or what we like to refer to as activating the Digital Thread.
In the information technologies (IT) world, we initially focused on back office and book keeping. This unlocked a whole new market category of “Systems of Record” applications emerged, based on key technology platforms such as relational databases. With the development of the Internet, it first hit people and created another breed of applications in a category referred to as “Systems of Engagement”, leveraging new platforms such as search, social graph, and messaging.
Asset-centric organizations have traditionally been focused on operational technologies (OT) and technologies that have evolved in parallel with IT. The rise of IoT is creating the conditions for convergence, yet we have not yet reached the same platform maturity. As we build outcome-based applications on top of digital twin platforms, we will also create a new market category, called “Systems of Asset.” Ultimately, true digital industrial transformation will happen at the intersection of IT and OT, and the digital twin will be the catalyst.