Frequently Asked Questions (FAQs)

Have questions about digital twins? You’re not alone. Here are some of the most common questions – and straight-up answers – drawn from the pages of Designing Digital Twins.

  • A digital twin is a virtual representation of a specific portion of the real world, designed with two-way interconnections that synchronize at defined frequencies and fidelities. Its purpose is to create measurable and repeatable changes in the real world, which can then be improved upon in the virtual space. Unlike static 3D models or simulations, a digital twin is inherently dynamic, strongly connected to real-world data, and focused on solving real-world problems rather than just visualizing or simulating them.

  • Digital twins excel at addressing cross-silo challenges because they integrate data, processes, and systems across traditionally isolated domains. In industrial organizations, silos such as maintenance, operations, or technology often operate independently, leading to inefficiencies when problems span multiple domains. Digital twins provide a framework to bridge these gaps by connecting real-world systems with virtual capabilities like AI, enabling actionable insights and coordinated solutions that improve outcomes across the entire ecosystem.

  • Digital twins often incorporate AI capabilities to analyze data, predict outcomes, and optimize processes. While AI provides probabilistic insights and advanced analytics, the digital twin serves as the ecosystem that connects these insights to actionable changes in the real world. Together, they enable organizations to design solutions that are not only technologically advanced but also directly tied to measurable improvements in physical assets, workflows, and decision-making.

  • Organizations often struggle with:

    • Siloed Thinking: Teams focus on their own domains without considering cross-functional impacts.

    • Weak Interconnections: Many so-called "twins" lack strong two-way connections between real and virtual worlds, reducing their effectiveness.

    • Overinvestment in Visualization: Excessive focus on high-fidelity 3D models can lead to "gold plating" without delivering ROI.

    • Underinvestment in use cases: Use cases can seem “too small” or “overwhelming in number” and as such, twin builders often focus on delivering Capabilities and hoping that valuable use cases will be built later.

    • Change Management: Resistance to new workflows or insufficient integration into existing processes can hinder adoption. To overcome these challenges, organizations must start with clearly defined problems and ensure their digital twin design aligns with measurable real-world outcomes.

  • The suitability of a problem for a digital twin depends on its complexity and the need for interconnected solutions. A valuable problem should:

    1. Require measurable change in the real world.

    2. Involve multiple systems or domains where traditional siloed approaches fall short.

    3. Justify the cost and complexity of building a twin by offering significant ROI or strategic benefits.

    By validating problems through frameworks like the Digital Twin Design Process (DTDP), organizations can ensure they focus on impactful solutions rather than chasing shiny technologies or low-value opportunities.

    1. Don’t try and define what you think a digital twin is.

    2. A great digital twin creates real change, measures that change, and improves that change. Aim everything at this.

    3. Decide if you’re trying to solve a real problem or trying to deploy new capabilities. Don’t try and do half of each or you’ll end up doing neither.

    4. If you’re solving a problem, work backwards from the problem to identify what needs to change in the real world, and then what technology you need to make that change.

    5. If you’re deploying capabilities, identify the key data types you use to operate every day and ensure your new capabilities can use these data types, and explicitly think about how this data gets from inside your twin to the real world to create change.

    6. Plan to build something small, and test how it works in the real world, and plan to add more data or extend to other area / assets. Budget to do this in stages, and don’t try and “one shot” it. Things will go wrong, but more importantly, you’ll learn a lot from doing this and you want to use those learnings to improve what you do.

    7. Don’t worry if people say your thing isn’t a digital twin, or doesn’t have enough AI in it. The purpose of a system is what it does – make sure it does something that changes the real world, and measures that change, and improves that change.