Digital twins are becoming more common across cities, infrastructure, construction, and asset management, but the term is still used loosely. A 3D model, a BIM environment, and a true digital twin are not the same thing, and that distinction matters when organizations are deciding whether the technology is worth the investment.

In a recent Net Zero Compare conversation, Hesham Gamal Gaafar, Manager of Digital Twin, Smart Cities and AEC at Esri Saudi Arabia, discussed how digital twins fit into the built environment, where they can create practical value, and why successful implementation depends on more than choosing the right platform. The conversation covered the relationship between BIM and GIS, the role of data quality and interoperability, sustainability applications, operational use cases, and the growing influence of AI.

🎥 Watch the Full Interview: The full interview explores how digital twins differ from BIM and GIS, where organizations are seeing the strongest return on investment, and how AI is being integrated into these systems. It also covers interoperability, data governance, implementation barriers, and practical sustainability use cases. Watching the full conversation provides more detail on where digital twins are already useful and where their limitations remain.

What Actually Makes a Digital Twin Different?

One of the central points in the discussion was that a digital twin should not be treated as another name for a 3D model. Hesham described the progression from CAD to GIS, BIM, city information models, and digital twins as a move toward richer and more connected representations of the physical world. CAD allowed teams to visualize geometry and spatial relationships. GIS added information to spatial features such as points, lines, and polygons, making it possible to understand assets and relationships at district, city, national, or global scale. BIM developed in a different direction, providing detailed, information-rich models of buildings and infrastructure components, including walls, floors, beams, materials, and systems.

What distinguishes a digital twin is the connection between the digital representation and the physical asset. Hesham explained that a digital twin requires a digital entity, a corresponding physical or target entity, and a link between the two that allows the digital version to reflect what is happening in the real world. That synchronization does not have to be instantaneous. Depending on the use case, updates may happen in real time, near real time, or at another agreed interval. Simulation, prediction, and advanced analytics can add value, but Hesham described them as additional capabilities rather than defining requirements.

The Strongest Business Case May Be in Operations

Hesham sees the operation phase as one of the strongest areas for digital twin return on investment because it is usually the longest part of an asset’s lifecycle. Planning, design, and construction may take months or years, but they eventually end, while buildings and infrastructure can remain in operation for decades. That creates more opportunities to use digital twins for energy optimization, natural resource management, maintenance, asset management, and other operational decisions.

Organizations should not wait until an asset is already operational before considering how a digital twin will be used. If the goal is to support operations, the necessary data structures, processes, and information flows should ideally be planned during design and construction. Building that foundation after a project is complete can require considerably more effort and cost.

Where Digital Twins Can Support Sustainability

Digital twins can support sustainability when they improve decisions rather than simply provide a more sophisticated visualization. During planning and design, digital modeling can support visual prototyping and material simulation, allowing teams to evaluate options before construction or manufacturing begins and potentially reduce reliance on physical prototypes. During operations, digital twins can support monitoring and control of energy-consuming systems, including lighting and air conditioning, based on how spaces are actually being used.

Hesham also gave the example of a district-scale water leak. If an organization has a digital representation that includes pipelines, valves, buildings, and the surrounding network, and that model is connected to current information from the physical system, it can help identify where the problem is occurring and which valves should be closed. The repair would still need to happen without a digital twin, but better information could reduce response time, limit water loss, and shorten the disruption for residents or businesses.

The environmental benefit in cases like this comes from better operational decisions. The digital twin provides the information needed to understand what is happening and respond more quickly.

Data Interoperability Remains a Major Challenge

Digital twins often depend on information from many sources, including BIM models, GIS data, sensors, Internet of Things systems, imagery, operational databases, and asset management platforms. Making those systems work together remains one of the most difficult parts of implementation. The challenge is not simply collecting more data, but ensuring that it is structured, compatible, current, and reliable enough to be used across teams and technologies.

Hesham described interoperability as a shared responsibility. Government entities can influence the market by encouraging or requiring open data formats. Standards bodies can develop common frameworks across areas such as BIM, GIS, and IoT, while software vendors need to support those standards in practice. Project owners also need clear requirements for data delivery, coordination, governance, and maintenance throughout the project lifecycle. Without that structure, organizations can end up with fragmented systems, incompatible formats, outdated information, and uncertainty over which version of the data should be trusted.

Why BIM and GIS Need to Work Together

Hesham also explained the different roles of BIM and GIS and why the distinction matters when organizations move beyond a single asset. BIM is well suited to understanding an individual building or infrastructure asset in detail because it can contain information about components, materials, dimensions, systems, and relationships between elements. GIS becomes more relevant when organizations need to understand how multiple assets relate to one another and to the wider environment.

A building is connected to roads, water networks, wastewater systems, energy infrastructure, transportation, neighboring buildings, and environmental conditions. For large developments and smart-city projects, combining BIM and GIS can therefore provide both detailed asset-level information and the wider spatial context needed for planning and operations.

Why Digital Twin Projects Fail

According to Hesham, digital twin projects often fail for organizational reasons rather than because the technology itself is incapable. Technology continues to improve quickly, but that does not solve unclear objectives, weak governance, poor data management, or a lack of ownership. Problems arise when organizations decide that they need a digital twin or an AI project before they have clearly defined what the system is supposed to improve.

Hesham argued that the sequence should be reversed. Organizations should first identify the decision, operational issue, or business process that needs improvement. From there, they can determine what information is required, how that information should be governed, which teams need to be involved, and which technology can support the solution. Even a technically capable system can underperform if teams do not know who owns the data, how it should be updated, or how the outputs are meant to influence business decisions.

Not Every Digital Twin Needs Real-Time Data

Another common assumption is that a digital twin must include continuous real-time sensor feeds. Hesham said this is not always necessary. Some applications require real-time or near-real-time updates, while others may work effectively with daily, weekly, or even monthly synchronization. The appropriate frequency should be determined by the business need and the decisions the system is intended to support.

This has direct implications for cost and complexity. If a weekly update is sufficient for a particular use case, building a real-time data architecture may add expense without creating meaningful additional value. A more advanced system is not automatically a more useful one.

What AI Can Add to Digital Twins

AI can extend the capabilities of digital twins when the underlying data is reliable enough to support it. Hesham highlighted predictive maintenance, anomaly detection, pattern recognition across sensor data, traffic optimization, scenario analysis, and infrastructure planning as potential applications. In predictive maintenance, for example, historical and current performance data can help identify patterns and allow an organization to intervene before an asset fails. In transportation, AI can analyze traffic behavior, identify recurring patterns, evaluate alternatives, or support decisions about new infrastructure.

These applications still depend on the quality of the information feeding them. Hesham summarized the problem with the familiar principle “garbage in, garbage out.” A sophisticated algorithm cannot compensate for unreliable or poorly structured data. AI can make a well-managed data environment more useful, but it does not remove the need to build that foundation first.

What Organizations Should Focus on Next

When asked what organizations should prioritize over the next few years, Hesham pointed to data governance and standards rather than the latest platform. Technology will continue to change, AI capabilities will improve, and new tools will enter the market. Organizations with clean, interoperable, well-governed data will be in a better position to adopt those technologies when there is a clear business case.

For companies starting from scratch, that can be an advantage because they have an opportunity to establish the right foundations instead of trying to repair poorly structured legacy systems. Hesham recommended starting with planning, problem definition, business requirements, and data standards before selecting a technology stack. That approach can also reduce unnecessary spending by helping organizations avoid tools and capabilities that do not contribute directly to the project’s objectives.

Technology Should Be Used Intentionally

Hesham closed the conversation with a broader point about technology adoption. Organizations should use technology intentionally rather than simply following trends, particularly as AI is being added to more digital systems. Technology requires computing infrastructure, energy, and other resources, so adding more of it without a clear purpose can create additional cost and environmental impact without delivering proportional value.

The relevant question is not whether an organization can add AI, more sensors, or a more sophisticated digital twin. It is whether those additions improve a defined decision, process, service, or outcome.

Conclusion

Digital twins can support better planning, operations, resource management, maintenance, and infrastructure decisions, but their value depends on how clearly the project is defined and how well the underlying information is managed. More sensors, more advanced visualization, or a more complex technology stack do not guarantee better results.

For sustainability teams and business decision-makers, the practical starting point is to define the problem, determine what information is needed, establish how that information will be managed, and then select the technology that fits the use case. That approach gives organizations a better chance of capturing real value from digital twins and AI without adding unnecessary cost or complexity.