Global data production is growing faster than ever before. By 2025, 463 exabytes (equivalent to the storage capacity of over a billion iPhones) will be generated each day. Contrary to popular belief, media and video platforms are not the main cause: enterprise data production is at a record high and will grow twice as fast as consumer data. Physical asset data is no exception to these developments.
The generous availability of data allows the creation of a Digital Twin: a realistic digital representation of a physical asset, with bidirectional synchronization between the two. The purpose of a Digital Twin is to better understand and control the real-world asset, to improve efficiency and decision-making. Additionally, new revenue streams can be generated, for example by selling the knowledge gained from a Digital Twin. Before a Digital Twin can be used to one’s advantage, an issue has to be tackled.
Over the life cycle of a physical asset, teams and organisations collaborate and share information. Between them, different specialized software applications are used. Because each application contains specific bits of information about a single asset, multiple partial representations of a single asset exist (Figure 1). When asset information within these applications is not interrelated in a meaningful way, it is nearly impossible to combine these representations into a single, complete Digital Twin.
Figure 1: Different representations of One Physical Asset
Take for instance the design, construction, and operation of a pyramid: An engineer at a design firm makes a 3D CAD design of the pyramid. Another organisation is responsible for verification and validation of the system against the client’s requirements, this is performed in a specialized MBSE (Model-Based Systems Engineering) application. During operation of the physical asset, maintenance is performed by a subcontractor who registers changes in their internal asset management applications. Sensors in the actual pyramid provide measurements about the actual state of the pyramid, like temperature, humidity and presence of people. The owner of the pyramid wants to have complete, correct and up-to-date insight on the pyramid’s state without needing to consult these applications. Integration of the asset and its different representations in applications are needed to be able to provide such insights. Without successful integration, the best case scenario is to have correct information available to all, but scattered across several applications. It is then impossible to answer questions that require a combination of information stored in these applications (e.g., what was the effect of last week’s replacement of air treatment installations of a specific type on average room temperatures?). As a result of scattered information, additional data becomes a burden rather than bringing new opportunities.
To have a complete Digital Twin of a physical asset, the missing relations between applications and real-world asset information like sensor measurements must be made. This requires technical compatibility between the different applications and systems containing representations about the asset: they must be able to handle the same data formats and be able to store or refer to information provided by other applications, however this is often not possible in applications built for a specific purpose. When multiple organisations are involved in a single asset, like in the example, technical alignment of applications across life cycle phases is even less likely. The reason for this is the freedom organisations have in picking their applications and vendors to suit their internal needs, combined with commercial incentives for vendors to lock organisations into their own software ecosystem. As a result, achieving technical alignment across the complete information landscape is uncommon.
Even when applications technologically align, shared understanding about the information contained by these applications is not guaranteed. Each team, individual, and organisation may have their own understanding (or lack thereof) about concepts. Without shared understanding, bringing information across is not possible. Imagine asking someone who doesn’t understand the concept of a pyramid to point you to the nearest one. While being able to hear your speech, he or she will fail to give a useful answer to your request.
In short: storing, finding and sharing complete and unambiguous information about physical assets is infrequently achieved. The main causes for this are:
Individuals, teams, and organisations lack a shared understanding of concepts. This complicates interpretation of information.
Technical incompatibilities between applications and the real-world asset restrict the possibility to create a single, complete representation from different partial asset representations.
Organisations lack control over the applications used by other organisations. This further increases the risk of technical misalignment.
To be able to create a complete and correct Digital Twin which can be used to share information within and between organisations, three things must be achieved:
Establish a shared understanding about concepts with those involved.
Example: The client requires a pyramid made from stone. The contractor prefers to use limestone. Those involved must know whether this fits the requirements.
Technically enable the creation of a single digital representation by allowing information from different partial representations to be combined into a single representation.
Example: Relating a stone block in a 3D model to real-world sensor measurements.
Enable organisations to share (parts of) this asset information with external organisations, regardless of the applications they use.
Example: The contractor shares detailed asset updates with the owner of the pyramid.
To establish a shared understanding between those involved, it must be possible for persons, organisations and applications to exchange knowledge in a meaningful and understandable manner. This process is divided into three sub-steps: Capture, Share and Apply.
Expert understanding about concepts usually resides in the heads of these experts, and each expert has their own area of expertise and view on reality. Therefore, in order to establish a shared understanding between experts, organisations and applications, this knowledge must first be captured in an information library. In a way, this ‘capturing’ is ‘releasing’ as well: knowledge used to be locked up inside the brain of an expert is now independently available. Because domain experts (e.g., pyramidologists) are usually not IT experts, capturing knowledge in such a way that it is reusable by others should be easy. The goal of this knowledge is to be able to refer others to it, hence the name reference data.
Before others are able to refer to our explanatory information, it must be shared. Logically, the only way to explain something is by using terms the other party understands. For instance, explaining what a pyramid is requires a shared understanding about triangle and construction. In case these concepts are unknown as well, we must resort to higher-level common understanding such as shape and corner. This approach of describing the meaning of things by describing the way in which they are related to other things, is called semantics.
Human language can be used to describe meaning, although in a slightly ambiguous way. Also, translations and sub-optimal computer interpretation are a burden for re-use in applications. A better way to share expert knowledge is by using Semantic Web standards such as Linked Data. These standards provide a way to express and share information in a formal and therefore computer-interpretable way by combining twenty-first century web technologies (like http://) and basic human language constructs (objects and relations). This guarantees a basic common understanding which can be used to exchange information, like explaining what a pyramid is. This enables effective communication between humans and computers, independent of the specific human languages or software applications used.
Those who are given access to shared knowledge may apply this knowledge to their benefit. For instance, to describe what is meant with a decimal value related to the line between the base and tip of a pyramid in 3D CAD software (its height). Such information is now given a formal definition and therefore no longer needs human interpretation. This provides a way to map asset information across partial representations. Additionally, it enables automatic verification of asset information against what we know to be true from the reference data.
Viewing a combined representation of a physical asset requires the combination of meaning, partial asset representations, and real-world information. Even when no application containing a partial representation supports such a combined representation, or when these applications are managed by different organisations. A Digital Twin platform can take care of this task, by integrating sources and adding relations between existing information. Reference data based on semantic web standards is one of the endless possible sources of information in this step.
Because the Digital Twin needs to provide a real-time accurate representation of the asset, it is important for integrations to be high-performance. To ensure reusability of configurations, the Digital Twin platform should offer a standardised (technology and information model) yet flexible (content and scope) interface based on commonly used IT standards.
Once a combined Digital Twin is constructed and available from a single entry point, humans can view and add information to the combined asset representation. This way, the Digital Twin representation can be completed by enriching it: Information not supported by source applications can be added directly in the Digital Twin platform. An example enrichment is adding manufacturer information for subsystems, while the source applications do not provide such information.
It is one thing to have a complete, meaningful and enriched Digital Twin available. Being able to hand this information over to external stakeholders is another. Asset owners increasingly demand standards like EN 17632 for Semantic Modelling and Linking because it offers compatibility and scalability benefits across organisations, projects and domains by enabling the reuse of knowledge and technology. Meanwhile, unreasonable requirements for contractors are avoided, like forcing the use of specific applications. Now, meaningful information from across applications can be combined into a single Digital Twin, and exchanged freely.
Figure 2 shows an example application landscape. Domain experts in the top left capture and share their knowledge using standardized and commonly known concepts. Their knowledge is applied in applications, including a Digital Twin platform. This platform is the backbone for combining and enriching partial asset representations and the real-world asset into a single digital representation. After handover, external parties are informed about the real-world asset through its Digital Twin representation.
The above solution requires tools. The sections below provide an overview of tools that can be used to create a landscape like shown above, regardless of what organisations are involved in and what software applications are used to manage partial representations of the asset.
Laces Suite is a SaaS (Software as a Service) suite that helps domain experts to capture and share knowledge with others using Linked Data technologies such as RESTful HTTP and RDF. Laces Libray Manager was developed to help domain experts in capturing their knowledge in a fast, easy and correct way. After capturing in the Library Manager, the shape of information can be altered easily and independently, for improved flexibility and future-proofing.
Domain experts are now a single click away from sharing their knowledge using Laces Hub. On this Linked Data Platform (containing a RESTful API for Linked Data), access to view and modify datasets can be managed per user and application. Those who are given access to information can now apply it to populate their software applications quickly and correctly, for instance by instantiating multiple doors at once. Aspects like cost, materials and dimensions are added automatically. Explicit references to the original source of other standards are kept, and can always be queried using SPARQL, allowing IT personnel to make complete and accurate configurations for domain experts and project staff.
Figure 3 provides an overview of the distinction between Capture, Share and Apply steps in Laces tooling.
Figure 3 - Laces Suite as a solution to establish shared understanding
Neanex Portal is a web-based application which acts as the central asset management platform in a Digital Twin application landscape. Because Neanex Portal focuses on information reuse and information handover, it is prepared to consume information that is published on a Linked Data Platform like Laces Hub to populate a Digital Twin and to add meaning to information originating from other sources. Through a standardized GraphQL API, bidirectional information exchange with partial representations is made possible. Furthermore, plugins for 3D CAD tools are available to automatically derive asset information from 3D drawings.
Within Neanex Portal, 3D and tree-based navigation of graph-based information allows users to know their asset inside out, without requiring every project member to use complex and expensive CAD software. The standardized API is pre-prepared for asset information handover in standards like EN 17632, RDF, IFC, CObie and more.
Figure 4 shows an application landscape where Neanex Portal is used to combine expert knowledge with partial asset representations from multiple sources. After enriching it using the web-based viewer, information handover is possible through its standardized but flexible API.
Figure 4 - Neanex Portal in a Digital Twin application landscape
If you recognize the challenges of meaningless, missing, outdated or scattered asset information within your organisation or in collaboration with other external stakeholders, take action now. In doing so, ensure future success by providing end-users with clear and easy tools that are prepared for future asset information requirements.
At Semmtech and Neanex, we have experience in setting up Digital Twins for information integration using open standards. As part of SPHERE, Neanex has provided a BIM-based Digital Twin Platform to optimise building life cycle management & energy efficiency. From SPHERE, the Building Digital Twin Association was founded to develop an open & ethical ecosystem for Building Digital Twins. Laces Suite is developed by Semmtech, who have worked on the European Road OTL for the European consortium of national road authorities (CEDR) and is an active member of NEN committee 391184 “Information-integration and interoperability”.
Together, Semmtech and Neanex are helping companies to make data the driver for their innovation. Besides software solutions for Digital Twins for Physical Assets, we offer end-to-end services from enterprise IT strategy and domain-specific consultancy to implementations and Linked Data publications. Please get in touch with us if you’d like to know more about the best next steps to become a data-driven organisation, or to simply get acquainted.