Among the Yoruba peoples of Nigeria, the firstborn of two twins is traditionally named Taiyewo, which translates as ‘slave to the second twin’.
A digital twin is a 3-D digital representation of a physical component, a real-world process or a fully functioning asset such as a plant or factory.
Like the firstborn Yoruba twin, the digital twin is often created during the design and conception stages of the physical counterpart, before the physical twin exists. And because its job over time is to mirror the changes in its physical counterpart as closely as technology will allow, the firstborn digital twin is indeed the slave in this relationship.
Growing a Digital Twin
The design stage uses CAD tools to create a digital twin in the form of a fairly generic 3-D model.
During construction or completion, the digital twin starts getting hooked up to IoT sensors that enable it to absorb simpler forms of data and knowledge from the real world.
However, it’s during commissioning or operations that the digital twin takes on a life of its own, through the power of AI and machine learning.
Properly implemented, in a collaborative environment, the digital twin is capable of delivering transformational leaps in operational efficiency and safety in its physical counterpart.
While the idea of a digital twin can be traced back to a paper published in 2002 by the University of Michigan, its technological roots lie in the design world of 3-D modelling and computer-aided design (CAD).
Consequently, in its simplest form, a digital twin is a 3-D representation of something physical. This model only describes the geometric form of the object, the x, y, z coordinates have been plotted in 3-D space, and it has probably been created using a CAD tool. Perhaps some renders have been applied to surfaces to make it look like the physical version.
Essentially, it’s a straightforward 3-D model but it is dumb in comparison to what it can communicate in later life.
The digital twin becomes more useful when it starts to absorb data and knowledge from its physical twin. Generally, the first way that this happens is through the addition of BIM (building information modelling) data to the model.
Once installed at the core of a large asset, for instance, the digital twin is able to accumulate even more data from sensors positioned in the asset, as well as input from inspection teams with tablets. In an ideal configuration, the digital twin will absorb every type of available information and store it in the form of a knowledge graph linked to a 3-D model. The digital twin is growing to become the principal font of reliable data for that asset, otherwise known as the single source of truth.
Coming of age
By harnessing the power of predictive analytics, the digital twin is able to analyse past and present data and draw useful insights from it. These insights can then be used to make informed predictions about the future of the asset which, once delivered to the teams on the ground, enable smarter decisions.
This changes the way people work. Equipped with mobile tools that stream real-time information and insights, people on the ground can do their jobs more efficiently than before, and make decisions based on the best and latest information. By empowering the teams that manage the plant, digital twinning opens up new ways of working and collaborating, even at a distance.
Expert support can be piped in remotely, reducing the need for on-site visits, because the asset’s digital twin is accessible for all to see, wherever they may be in the world.
With the aid of augmented reality devices, technicians at work can see the virtual asset super-imposed on the real one. This improves orientation and can alert them to hidden hazards. For example, maintenance teams can quickly identify what kind of fluid an insulated pipe contains, be it a hydrocarbon or an acid, and be made aware if it is at a high temperature or at high pressure.
Once the technology is fully embraced, the whole focus and structure of the business can change, eliminating old roles, creating new ones, and potentially delivering whole new streams of revenue.
In other words, digital twins drive innovation.
With advances in artificial intelligence and machine learning, otherwise known as predictive analytics, digital twins are already capable of predicting the future of their physical sibling. They do this by analysing data and using the insights they’ve gleaned to anticipate problems or issues before they actually happen. This has big implications for the efficiency of maintenance programmes, and promises to deliver better decision making in general.
All the while the digital twin is storing and organizing ‘lessons learned’, which in aggregate become a valuable repository of knowledge. Because the repository is owned by the operator of the plant or factory, everyone is contractually obliged to use the platform and share their knowledge.
This means that contractors, who previously had no incentive to share their data or know-how, are prevented from hoarding their valuable knowledge. This puts the owner back in control of their asset and allows them to manage risk better, because they have all the information they need at their fingertips.
So, as the digital twin gathers knowledge and data, it rapidly becomes indispensable. Having shadowed the physical asset over the full course of its lifecycle, the value of that knowledge is retained even after the physical plant has been de-commissioned. In other words, the digital twin has the potential to outlive its physical sibling.
In Yoruba culture, if one twin dies the parents commission a wooden sculpture to represent the child, an Ibeji, which they then look after as if it was a real person.
In the world of the asset, the physical twin is quickly forgotten after de-commissioning. But the data lives on as digital wisdom, ready to shape decisions about the next generation of assets.