Advances in reality capture and artificial intelligence are accelerating the use of digital twins beyond traditional engineering domains. Virtual replicas of physical processes or systems can now simulate more complex, dynamic environments, supporting capital allocation, strategic choices and operational efficiency in ways that were previously difficult to model. Underpinning technologies are becoming mainstream across industries, with high levels of cloud adoption and widespread deployment of Internet of Things (IoT) platforms, alongside growing use of extended reality (XR). Market projections indicate rapid expansion as organisations explore new sources of value, from faster design cycles to more resilient operations. Healthcare, government, financial services and consumer sectors are beginning to leverage these capabilities to solve problems that demand rich spatial data, realistic simulation and timely decisions.
From Closed Systems to Open Environments
For decades, digital twins were largely associated with contained, deterministic settings such as factory floors or engineered components. Manufacturers used sensor-rich twins to simulate production lines, discover bottlenecks and reduce costs. Engineers in precision environments, including motor racing, modelled parts to test performance thresholds before deployment. These applications benefited from stable boundaries and well-understood data flows.
The technological landscape is now shifting. Greater accessibility of reality capture through drones, lidar and high-fidelity imaging, combined with more sophisticated AI techniques, is enabling twins to represent open, variable environments. This broadens the remit from plant-level optimisation to scenarios that involve uncertainty and change, such as market responses to product launches or integration choices during mergers. Simulation methods including Monte Carlo, agent-based and discrete event approaches are paired with twins to explore outcomes and stress test decisions. The result is a move from mostly operational gains toward strategic decision support, with organisations seeking both efficiency improvements and new growth opportunities executed in cost-effective, worker-friendly ways.
Must Read: Precision Care Through Medical Digital Twins
Adoption drivers are visible in cross-industry data. Survey findings show 84% of respondents using cloud computing, 72% using IoT sensors, devices and platforms and 26% already deploying XR or augmented and virtual reality, with a further 26% planning deployment within three years. Projections for the global digital twin market suggest growth from nearly €12.1 billion in 2023 to €240.9 billion by 2032, reflecting the expanding scope of applications and the diffusion of enabling technologies across sectors.
New Spatial Data and Synthetic Data Close Gaps
Open environments present data capture challenges. Bridges spanning rivers, urban corridors or rugged terrain, as well as minuscule and dynamic areas within the human body, are difficult to instrument continuously. Spatial computing offers a way forward by bringing together operational telemetry, enterprise data, reality capture, geospatial and temporal information and three-dimensional models to mirror the physical world in virtual form.
Illustrative examples show how this can compress time and enhance accuracy. In public infrastructure, a city turned to drones with real-time onboard analysis to accelerate inspection of bridges that carry heavy daily traffic or date back more than a century. What previously took months was reduced to minutes, with issues identified rapidly for review in web or virtual reality environments and converted into work orders. In life sciences and healthcare, clinicians are experimenting with personalised digital twins for epilepsy management that combine imaging with implanted, Bluetooth-enabled devices. Bidirectional data flows support real-time monitoring and predictive adjustments, with embedded electrodes interacting through IoT to continually refine therapeutic parameters. These approaches highlight how spatial computing and in vivo data streams can underpin patient-specific models in areas where stimuli are not always predictable.
Where real-world data is sparse or impractical to capture, synthetic data can fill crucial gaps. An energy provider facing limited defect imagery across a sprawling power grid generated more than 2,000 3D synthetic images to train computer vision models. The resulting models improved defect detection by 67%, reducing asset downtime and improving customer experience. In technology and mobility, a partnership between Nvidia and Uber combines AI training infrastructure with extensive trip data to support the development of autonomous vehicles. Highly detailed synthetic datasets can augment real-world driving records to inform complex models and scale safer testing regimes. Together, spatial and synthetic data approaches enable richer, more resilient simulations when fully instrumented reality is unattainable.
Design for Human Decisions and Operations
Human behaviour introduces variability that is difficult to track and, in many contexts, inappropriate to monitor closely due to privacy considerations. Rather than seeking full behavioural capture, organisations are applying digital twins to improve the design of spaces and flows so that operations adapt more smoothly to human decisions.
In consumer services, a quick service restaurant (QSR) chain uses a digital twin to replicate end-to-end fulfilment and evaluate operational scenarios across ordering, payment and service flows. A 3D visualisation of the restaurant enables the team to test alternative layouts, examine the impact of customer traffic patterns and align operational processes with digital and marketing choices. The aim is to identify high-leverage changes that raise accuracy and timeliness while containing costs.
In financial services, large-scale transformation can be accelerated by virtualising physical estates. When BMO acquired more than 500 branches in the purchase of Bank of the West, the bank used 3D capture technology from a digital twin provider to create virtual replicas of the new locations. The twin environment supported simulation of layouts, branding elements, ATM placement and signage, allowing redesign to proceed without dispatching survey teams to every site. The approach saved hundreds of thousands in travel expenses and more than 6,000 hours of survey work, demonstrating how virtual models can streamline integration efforts and standardise execution at scale.
These semi-open system strategies prioritise operational flow, inventory and navigation rather than full behavioural modelling. By focusing on the environment that shapes decisions, rather than exhaustive tracking of the decisions themselves, teams can deliver measurable gains while respecting constraints around privacy and practicality.
Digital twins are moving from stable, sensor-dense settings into complex environments that demand richer data and more adaptable simulation. Leaders aiming to capture value from these capabilities can start by identifying where decision efficiency or innovation speed lags, building talent pipelines through partnerships that extend beyond organisational boundaries and designing data architectures that support multimodal spatial inputs. Evidence from public infrastructure, healthcare, energy, mobility, retail and banking shows how spatial computing, synthetic data and process-centred design can compress time, improve accuracy and reduce costs. As adoption grows, organisations that integrate these practices are positioned to build more resilient operations, support better decisions and prepare for future uncertainty across healthcare and other sectors.
Source: Deloitte
Image Credit: iStock
