Analysing the Value of Digital Twinning Opportunities in Infrastructure Asset Management

Abstract

Many studies and technology companies highlight the actual or potential value of Digital Twins, but they often fail to demonstrate this value or how it can be realised. This gap constitutes a barrier for infrastructure asset management organisations in their attempt to innovate and incorporate digital twinning opportunities in their decision-making processes and their asset management planning activities. Asset management planning activities often make use of existing value-based decision-support tools to select and prioritise investments in physical assets. However, these tools were not originally designed to consider digital twinning investments that also compete for funding. This paper addresses this gap and proposes a value-based analysis for digital twinning opportunities in infrastructure asset management. The proposed analysis method is tested with three rail and road infrastructure case studies: (i) real-time monitoring of a power transformer; (ii) BIM for the design, construction, and maintenance of a new railway line; and (iii) infrastructure displacement monitoring using satellite data (InSAR). The study shows that the proposed method provides a conceptual construct and a common language that facilitates the communication of digital twinning opportunities in terms of their relevance in different contexts. The proposed method can be used to support the investment decision-making process for investments in both physical and non-physical assets and help derive maximum value from the limited available resources.

1. Introduction

In infrastructure asset management organisations, resources are often scarce, so asset managers are pushed to allocate the available resources to opportunities that deliver the highest value to relevant interested parties. However, understanding and collectively agreeing on what constitutes value is challenging. This results from a combination of factors, namely [1]:

• Several interested parties are involved in decision-making, sometimes with conflicting objectives;
• There are numerous options available, and information on how they contribute to the objectives is often limited;
• When many units/assets share resources, the optimal resource allocation to each unit/asset hardly ever represents the best collective resource allocation—“The whole is greater than the sum of the parts”;
• Decisions often produce second- and higher-order effects in complex systems like infrastructure asset systems. These effects are difficult to measure and may affect the types of impacts and their real magnitude.

As proposed by some authors [2,3], several perspectives exist on the value construct. The six-capitals model is one of the most fundamental perspectives on value reporting [4]. This model contains six forms of capital (or value domains, as called by [3]) an organisation uses or affects in the world: financial, manufactured, intellectual, human, social and relationship, and natural. The six-capitals model is, however, a high-level approach to value, which needs to be more detailed from a management perspective [3].

Asset management is the coordinated activity of an organisation to realise the most value from assets [5]. According to ISO 55000 [5], value is the result of satisfying needs and expectations regarding the existing stakeholders. Therefore, what constitutes value will depend significantly on each organisation and the corresponding stakeholders (such as customers, investors, employees, regulators, and local society). Value realisation normally involves balancing costs, risks, opportunities, and performance benefits [5].

Regarding the global digital twinning trend in infrastructure asset management, many technology companies (representing the “supply” side) highlight the current or potential value of “Digital Twins” (DTs). However, they do not detail this value or how it can be achieved. This research gap was identified in a previous literature review [6], and it constitutes a barrier for infrastructure asset management organisations (the “demand” side) in their attempt to innovate and incorporate digital technologies in their processes and assess what the market can really offer to help them achieve their organisational goals.

For example, Singh et al. [7] argued that there is a lack of consensus regarding the value of DTs. The authors claimed that understanding the value DTs can bring to the business is crucial so that the correct type of DT (according to the different characteristics and types—see Section 2.3) can be chosen and provide maximum profits. Dirnfeld [8] also claimed that deciding the best approach to implement DTs is challenging, so adequate guidelines are needed. Sanfilippo et al. [9] argued that there is a DT implementation gap in the field of road infrastructure, with few examples showing the value of adopting DTs. Yan et al. [10] stressed that future research should focus on developing standardised methodologies, frameworks, and guidelines for implementing DT in this sector. This research gap became so clear that the World Road Association [11] set a group of recommendations to follow on this subject, namely to support the development and implementation of DT-related initiatives within the road sector.

When applying the asset management approach to the digital twinning of physical assets, particular aspects should be considered. For example, replicating physical assets in the digital space promotes acquiring or creating new assets. These new assets can be physical (or tangible, such as IoT infrastructure, sensors, and data centres) or digital (intangible, such as data, models, licences, and software), and both should be managed according to the asset management principles. Extracting the maximum value from the asset portfolio will involve not only an alignment between the organisational functions but also between the different types of assets within the asset portfolio.

As illustrated in Figure 1, each organisational level deals with specific decisions. Managerial and executive levels tend to decide more on strategy and planning, whereas the operational levels focus more on preparation and operations. These decisions depend on information requirements, which are linked to Information Quality Levels (IQL) [12]. Lower IQL (operational levels) are data-intensive, require more detailed data, and are more suitable for automated data collection methods. On the other hand, higher IQL (strategic levels) need information in very summarised formats, which tend to be associated with manual data collection and processing methods [12]. Understanding these differences can help asset managers map the organisational information needs and, by comparing them to their current context, identify digital twinning opportunities with the potential to be analysed and implemented [13].

Asset management decisions vary greatly in complexity, so it is inappropriate to apply the same level of sophistication to all decisions [16]. Simple, low-impact, and operational (e.g., asset-centric) decisions make use of informed and pre-determined rules and guidelines (“business as usual”). More complex and higher-impact decisions (e.g., with multiple influences, inter-dependencies, or stakeholder interests) require specific tools and more systematic, rigorous, and auditable decision-making processes [16]. Value-based decisions are primarily linked to managerial levels. However, they also relate to executive levels (see Figure 1) as they typically occur on a longer time basis (e.g., years) and address complex choices regarding capital projects, investment budgets, and asset management plans, among others [14]. According to the IAM [16], value optimisation, although applicable to different asset portfolio levels, is more suitable for system/network and portfolio levels. Nevertheless, the IAM also stresses that asset-level decisions must consider the asset’s contribution at the system level [16].

The literature mentions value-based decision-support tools. Some of these tools include academic contributions (e.g., [17,18]), investment planning solutions developed by the technology market (e.g., Asset Investment Planning software from Copperleaf Technologies, ASSETSVALUE from Assetsman), and industry practices such as the ones published in [19]. However, these tools are usually focused on analysing and evaluating infrastructure investments, leaving a significant gap in terms of digital twinning investments and the consideration of data assets as a type of asset with increasing relevance in asset management.

Considering the challenges presented above, this paper aims to analyse the value of digital twinning opportunities according to asset management principles. The authors propose a value-based analysis method to achieve this goal. This analysis method provides a structure of thought and a common language to facilitate the communication [20] of digital twinning opportunities in terms of their different contexts and impacts. It may be used to support the investment decision-making process, considering the tight competition for investment and the need to extract the best value from the limited available resources.

The present article is structured as follows: Section 2 presents the case study demonstrators used to validate the proposed method of analysis; Section 3 presents the value-based analysis method; Section 4 presents the results obtained from a demonstration of its application to three case studies from the rail and road infrastructure domain; and the paper concludes with Section 5, dedicated to the main research outcomes and possible future developments.

2. Case Study Research

2.1. Research Approach Supported by Case Studies

Although this work focuses on road and rail infrastructure networks, the value-based analysis method proposed here was designed to be applicable in other scenarios using the necessary adjustments (e.g., context, value framework, criteria).

Constructing and validating an analysis process for such a broad scope is challenging. A hypothetico-deductive research approach was used to achieve this. The hypothetico-deductive method uses an existing research gap and the formulation of an initial hypothesis to build a theoretical model that will be tested, improved, and corroborated with real-world data [21]. This method uses a continuous cycle of inductive (“from the world to the mind”) and deductive (“from the mind to the world”) research streams during model development [22].

The inductive stream aims to establish conclusions with a more generic scope from conclusions verified at a more particular level. So, it is usually based on experience [21,22,23,24], namely by using case studies to validate the proposed model. According to Neuman [25], case studies enable a connection between the micro-level (actions of individuals) and the macro-level (large-scale structures and processes). Case study research provides an extensive list of strengths, some of them closely related to the challenges derived from this particular application: conceptual validity; heuristic impact (i.e., providing further learning, discovery, or problem-solving); identification of causal mechanisms (make visible mechanisms by which one factor affects others); ability to capture complexity and trace processes; calibration (enables the adjustment of abstract concepts to real-life experiences, problems, and widely accepted evidence); and holistic elaboration (allows to elaborate holistically on an entire situation or process and to incorporate multiple viewpoints).

The next section presents the case study demonstrator used in this work.

2.2. Rail and Road Case Studies in Infraestruturas de Portugal, S.A.

Infraestruturas de Portugal, S.A. (IP) is a public body responsible for developing, operating, and maintaining road and railway networks in Portugal. This organisation manages almost 15,000 km of national roads and 3000 km of railway networks [26], which are distributed by multiple asset groups.

IP identifies the power of information and technological innovation as two of the five major trends in the transportation sector. Within these subjects, IP highlights subtopics such as Big Data, Artificial Intelligence, automation, connectivity, and new infrastructure monitoring processes [27].

In 2023, IP mapped fifty innovation challenges [27], one related to creating road and rail infrastructure DTs. While IP recognises DTs as innovative solutions, their full meaning and potential impacts still need to be understood. To bridge this gap, experts from IP were challenged to identify a list of potential digital twinning opportunities based on their knowledge, ongoing projects, and the innovation challenges previously mentioned (

Table 1. Digital twinning opportunities identified for the rail and road infrastructure networks.

Case Study	Network	Brief Description
[A]	Rail	Real-time Dissolved Gas Analysis system in a power transformer at Salreu traction substation, located on the Linha do Norte railway line
[B]	Rail	Dynamic train weighting triggered by trains’ motion, applied to a track segment in the Linha do Norte railway line
[C]	Rail	BIM-supported fatigue life prediction system for the Várzeas railway bridge
[D]	Rail	Landslide and rockfall detection system, applied to a track segment on the Douro railway line
[E]	Rail	Automatic structural health monitoring of the walls of the Rossio railway tunnel
[F]	Rail	BIM methodology for the design, construction, maintenance, and renewal of the new high-speed rail line connecting Lisbon and Porto
[G]	Rail	Adoption of rail control software with signalling design automation and verification for the new Lisboa—Santa Apolónia railway station
[H]	Rail	Development of a Digital Shield for the rail track platform to improve safety during construction works around the rail corridor and reduce the impact on operations
[I]	Road	Real-time monitoring system of pavement performance in a road section of the IC5
[J]	Road	Monitoring of pavement condition using a vehicle equipped with a Laser Profiler that allows obtaining three-dimensional cross-sectional profiles
[K]	Rail and Road	Displacement monitoring of rail and road infrastructures (earthwork structures and bridges) using satellite data (InSAR)

).

Then, IP experts were asked to select three distinct opportunities with sufficient technological maturity—according to NASA’s Technology Readiness Levels [28]—to be eligible for implementation and used to validate the applicability of the analysis methodology. The chosen opportunities are presented in

Table 2. Digital twinning opportunities selected to validate the value-based analysis method.

Case Study		Network	Description
[A]	[29]	Rail	Real-time Dissolved Gas Analysis (DGA) system in a power transformer at the Salreu traction substation, located on the Linha do Norte railway line. The Hydran M2-X unit continuously monitors and communicates gas and moisture levels dissolved in the power transformer insulating oil.
[F]	([30], for illustration purposes)	Rail	Adoption of BIM methodology for the design, construction, maintenance, and renewal of the new high-speed rail line connecting Lisbon and Porto.
[K]	[31]	Rail/Road	Displacement monitoring of earthwork and engineering infrastructure using satellite data. Interferometric Synthetic Aperture Radar (InSAR)—a remote sensing technology based on radar waves—is used to detect ground or structural motion on a network-wide scale.

. They are used first to validate the proposed UNI-TWIN assessment model (see Section 2.3) and then to validate the proposed value-based analysis method of digital twinning opportunities (see Section 4).

2.3. UNI-TWIN Model

Digital twinning is a continuous and variable process of representing physical assets in the digital space across a set of dimensions with various levels of complexity [32]. The unified assessment model for the Levels of Digital Twinning (LoDT)—entitled UNI-TWIN ([32])—is a comprehensive and unified representation of the many dimensions embodied in the concept of digital twinning of physical assets. It supports digital twinning strategies within organisations by continuously evaluating the current situation, comparing it with other contexts (internal or external), and identifying potential development opportunities. UNI-TWIN is not a maturity assessment model wherein higher LoDT do not necessarily imply greater maturity. Rather than assessing maturity, the UNI-TWIN model provides a comprehensive representation of the complexity in current or possible scenarios based on the dimensions constituting the overarching concept of digital twinning.

Table 3. Summary of UNI-TWIN dimensions (adapted from [32]).

Dimension	Description
Hierarchy	The hierarchical level of the physical assets
Connection	The type of data connection between the physical and digital spaces
Synchronisation	The frequency at which data are integrated into the digital space
Geometric representation	The type of geometric representation of the physical space
Non-geometric representation	The level of representation of non-geometric characteristics of the physical space that are relevant to the defined purpose
Intelligence	The type of intelligence associated with data analysis
Interface	How users interact with the information generated in the digital space
Accessibility	The scope of users that access the information generated by the digital space
Autonomy	The autonomy level of digital space in decision-making

summarises those dimensions, while

Table 4. Proposed UNI-TWIN model for assessing the LoDT (adapted from [32]).

Level	Hierarchy	Connection	Synchron.	Geom. Rep.	NGeom. Rep.	Intelligence	Interface	Accessib.	Autonomy
1	Asset subcomponent	No connection	No synchronisation	No geometric represent.	No representation of non-geometric characteristics	Descriptive	No interface	Single user	No autonomy
2	Asset component	Manual data flow in both ways	Monthly/yearly	Conceptual	Necessary non-geometric requirements, with flaws	Diagnostic	Local access	Minimal	User- assistance
3	Asset	Semi-automatic in one way	Daily/weekly	Approximate	Necessary requirements without major data flaws	Predictive	De-centralised and shared access	Limited	Partial autonomy
4	Asset system/group/class	Automatic in one way	Hourly/minutes	Precise	Necessary and important requirements. Some data flaws can exist	Prescriptive	Immersive	Advanced	High autonomy
5+	System of Systems (portfolio)	Automatic in both ways	≤seconds/event-driven	As-built	All non-geometric requirements	Cognitive	Smart hybrid	Full	Limited options for humans

provides the complete assessment model (a more detailed version can be found in [32]).

2.4. Application of the UNI-TWIN Model to the Case Studies

Table 5. LoDT for rail and road case studies.

Case Study	Hierarchy	Connection	Synchron.	Geom. Rep.	NGeom. Rep.	Intelligence	Interface	Accessib.	Autonomy
[A]	(1)	(4)	(5+)	(1)	(3)	(1)	(3)	(1)	(2)
	Asset subcomponent (oil of a power transformer unit)	Automatic from physical to digital space (via Ethernet)	Readings sent to user every 15 s	No representation is needed for this phase	PT ID, location, temperature, moisture, and gas readings; it lacks gas characterisation	Descriptive (for now, data are analysed descriptively)	Decentralised access to data (IP server)	A single user has access to these data	Sends alerts in case of abnormal values
[F]	(4)	(2)	(3)	(5+)	(3)	(1)	(3)	(5+)	(1)
	New rail line	Inputs of internal and external data are mainly manual/off-line	Data inputs are most frequent during construction (days)	As-is BIM model, LOD 300	Necessary requirements are represented with adequate quality, quantity, and granularity without major data flaws	Descriptive (clashes, quantities, etc.)	Decentralised and shared access	Multiple users across organisations	No autonomy
[K]	(4)	(4)	(3)	(3)	(4)	(3)	(3)	(3)	(2)
	Group of 18 earthwork and engineering assets	Automatic from physical to digital space (via satellite)	Sentinel-1 satellites allow radar image acquisition every 12 days	The elements are represented in GIS with their approximate quantity, size, and shape.	Necessary and important requirements (vertical and horizontal displacements, displacement risk)	Historical analysis and predictive tools for exceeded safety thresholds forecast	Online visualisation tools (shared access)	Multiple users across departments	Automated alerts once safety thresholds are crossed

briefly describes the case studies and their characteristics according to the UNI-TWIN model. Each description is based on available documents and project information provided by IP experts.

Based on the characteristics of each case study and the levels assigned to each dimension of digital twinning, it is possible to construct the LoDT radar shown in Figure 2.

As displayed in Figure 2, the power transformer case study ([A]) shows higher LoDT in the components of “connection” and “synchronisation”. This is justified by the case study’s purpose of real-time monitoring. Case study [A] presents the lowest LoDT in terms of “hierarchy” (level 1) since it is focused on the insulating oil of a power transformer located in a traction substation. This case study aims to develop predictive capabilities based on collected data but currently only has the capacity for descriptive analyses (level 1 in “intelligence”).

The adoption of BIM for the new high-speed rail line ([F]) exhibits the highest LoDT in terms of “geometric representation” and “accessibility”, indicating some expertise in these dimensions due to its collaborative and visualisation capabilities.

Case study [K] stands out by showing the most balanced group of digital twinning dimensions in terms of LoDT. It is also the case study with the highest level in terms of “intelligence” (due to its goal of forecasting long-term displacements) and of “non-geometric representation” (by adding relevant non-geometrical information, such as displacement risk).

The dimension with the lowest level of development is “autonomy”, which, at most, reaches level 2 (case studies [A] and [K]). This is generally attributed to the fact that most of the digital twinning opportunities presented here aim to provide better support (by improving data quality and timeliness) to asset management decisions taken by humans and not to replace their actions.

3. Value-Based Analysis of Digital Twinning Opportunities

3.1. Method of Analysis

Digital twinning is present in multiple environments, differing in types of physical assets, digital twinning dimensions, and the corresponding levels of complexity (LoDT).

The method for analysing the value derived from digital twinning opportunities relies on the conceptual foundations of value laid down by ISO 55000 [5]. Although applied to digital twinning opportunities, this analysis method is grounded on the concept of value taken in its broader sense, so its structure is conceptually prepared for application to other types of opportunities and organisational contexts using the necessary adjustments (e.g., context, value framework, criteria). This analysis method provides input to opportunity evaluation, to decisions on whether opportunities should be implemented and how, and on the most appropriate implementation strategy and methods [33].

The proposed analysis method follows the principles of multi-criteria value measurement. This method analyses alternatives (opportunities) and their different impacts on several criteria based on qualitative value judgements. It accommodates the uncertainty associated with those judgements and the multiple and conflicting objectives that typically exist in asset management (costs, benefits, and risks) through a comprehensive and aligned value framework [34].

Figure 3 illustrates the steps of the analysis.

The proposed analysis method initiates with the construction of the value framework based on the needs and expectations of stakeholders (3.2); it proceeds with choosing the analysis criteria (3.3), analysing the context (4.1) and the impacts of each opportunity (4.2 and 4.3); ending with the overall value analysis of each opportunity (4.4).

Data inputs were collected from IP documents and other relevant sources and through individual semi-structured interviews with experts from IP, following the best practices regarding interview-based collection methods (e.g., [35]). The sample of participants in the interviews was selected considering their particular professional experiences with the case studies, allowing for the deepening of the analysis and the probing into cross-related issues during the interviews. The interviews were performed as part of a routine of continual improvement of the method of analysis. The first iterations assisted the design and development of the analysis method until the final consolidated version of the proposed method was achieved. The interviews did not follow a pre-defined script but instead focused on the structure of the analysis method (Figure 3).

3.2. Value Framework

To grasp the value derived from digital twinning opportunities, the needs and expectations of internal and external stakeholders need to be identified first.

IP manages a portfolio of critical assets and asset systems that are quite diverse and distributed throughout Portuguese territory. These circumstances mean that IP has a vast network of stakeholders with competing interests [36]. Considering the context of IP and some internal documents (namely the Strategic Asset Management Plan [37]), the needs and expectations of the most relevant internal and external stakeholders were identified (

Table 6. Needs and expectations of relevant IP stakeholders.

Stakeholder Group	Network	Relevant Stakeholders	Needs and Expectations
Shareholder (external)	Rail, Road	Portuguese State	Sustainable construction Sustainable mobility Efficient management (quality vs. cost) Good reputation Public service Rationality and criteria in investment selection Reduction in accidents
Users (external)	Road	Private or collective users of the national road network	Good condition and functionality Safety Accessibility Network availability according to service levels Reduced costs
		Concession holder	Collaboration in contractual relationship Control of contractual obligations
	Rail	Rail operators, rail service users	Fair service pricing Information Availability, punctuality, reliability Safety Reduced costs
Regulatory body (external)	Road	AMT *, IMT *, ANSR *	Compliance with the Concession Contract
	Rail	AMT *, IMT *	Compliance with the Contract Programme Compliance with safety requirements
Local bodies (external)	Rail, Road	Municipalities, CCDR *, neighbouring municipalities	Equity and transparency Accessibility Information
Suppliers (external)	Rail, Road	Other concession holders, toll operators, design and construction companies, service providers, suppliers	Compliance with contractual obligations Equity and transparency
Labour organisations (external)	Rail, Road	ACT *	Compliance with legislation
Media (external)	Rail, Road	Media	Quick, accurate, and up-to-date information
Workers (internal)	Rail, Road	Workers’ committees, labour unions	Safety Training Fair pay Prospect of career progression

* AMT: Authority for Mobility and Ground Transports; IMT: Institute for Mobility and Ground Transports; ANSR: National Road Safety Authority; CCDR: Commissions for Regional Development and Coordination; ACT: Authority for Working Conditions.

).

As stated by [5], realising value normally involves balancing costs, risks, opportunities, and performance benefits. In that regard, the value framework can be constructed from the combination of impacts on the assets—in terms of asset cost, risk, and performance—and the impacts on relevant stakeholders.

The context of each opportunity is another relevant aspect when constructing the value framework. The investment required to implement a given opportunity or the availability of external funding (e.g., through European financing) is an example of contextual aspects that affect the overall value of that opportunity. In this respect, a group of contextual aspects are integrated into the value framework.

Internal and external documents regarding value and multi-criteria decision-making in road and rail networks were also used to help build the value framework. The final value framework is presented in Figure 4 as a value tree formed by the contextual aspects and the impacts of each opportunity. Table 7 describes each value criterion of the value framework and provides examples of metrics that can be used to analyse impacts.

Value frameworks can be adapted to different asset management contexts [38]. Different organisations may arrange and organise their value framework criteria in specific ways, matching them with different categories and subcategories in a more or less aggregated manner, for example. Value criteria can sometimes have dependency relationships (e.g., safety and asset risk) or be grouped according to multiple perspectives. For example, the operational expenditure of an asset (OPEX) could also be included within the financial perspective of asset performance. Another common issue arises when addressing risk. Defined as the effect of uncertainty on objectives, risk can have different aspects and be applied at different levels (strategic, operational, programme, project, etc.) [33]. In the context of this work and regarding asset scope, risk is analysed from the perspective of asset failure, i.e., the effects of uncertainty related to asset failure. Asset risk is here decomposed into failure probability (probability of an event occurring) and failure consequence (consequences if the event occurs, such as temporary speed restrictions, corrective actions needed, accidents, etc.), following a Bow tie-type of analysis [39]. At the stakeholder level, the risk is distributed across various dimensions (e.g., safety, corporate image, utilisation) since, at this level, the sources of risk go beyond the asset context and may depend on other (internal and external) factors, such as human resources, users, customers, suppliers, economic context, etc.

Table 7. Description of the value framework criteria (abbreviations).

Description
Context	Hierar	Level of asset hierarchy to which the opportunity is referring
	LoS	Level of service associated with the asset(s), expressed in terms of network level
	Subcl	Asset subclass(es) to which the opportunity refers
	Invest	Upfront investment needed to implement the opportunity
	Financ	Type of financing opportunities related to the investment needed
	Synerg	Type of synergies associated with the opportunity in hands
	Compl	Type of commitments and requirements (political, legal, local/national/international) associated with the opportunity
Asset(s)	CAPEX	Funds used to acquire or upgrade long-term physical assets
	OPEX	Ongoing costs that the organisation incurs as a result of performing its normal business operations
	Avail	Time for which the asset is available to be operated relative to the total time. It includes relevant aspects such as reliability and service punctuality (affected by infrastructure).
	Capac	Ability of the asset(s) to support certain operational requirements (volume, speed, size, load, etc.)
	Obsol	Process of assets becoming antiquated in comparison with newer versions
	Environ	Environmental externalities derived from the asset life cycle activities
	Condit	Level of asset integrity, affecting the probability of failure
	Conseq	Consequences derived from asset failure
Relevant stakeholders	Safety	Protection of passengers, workers, and other people from physical harm or security breaches
	Utilis	Demand for services provided by the assets
	Know 1	Ability to use information in decisions and actions, and the impacts on process efficiency and effectiveness
	HR	Development of workers’ capabilities, career, and job satisfaction
	Access	Capacity of users to access the service (accessibility) and the ability of infrastructure to connect with other assets/services/systems (interoperability)
	Image	Overall corporate image, communication, and transparency

¹ This criterion includes the impacts on knowledge and process efficiency. This was decided based on two reasons: simplification of the value framework; the two dimensions showed high correlations with the topic “asset knowledge enablers”, demonstrated in the report “Criteria for evaluating asset management indicators” [40].

Table 7. Description of the value framework criteria (abbreviations).

Description
Context	Hierar	Level of asset hierarchy to which the opportunity is referring
	LoS	Level of service associated with the asset(s), expressed in terms of network level
	Subcl	Asset subclass(es) to which the opportunity refers
	Invest	Upfront investment needed to implement the opportunity
	Financ	Type of financing opportunities related to the investment needed
	Synerg	Type of synergies associated with the opportunity in hands
	Compl	Type of commitments and requirements (political, legal, local/national/international) associated with the opportunity
Asset(s)	CAPEX	Funds used to acquire or upgrade long-term physical assets
	OPEX	Ongoing costs that the organisation incurs as a result of performing its normal business operations
	Avail	Time for which the asset is available to be operated relative to the total time. It includes relevant aspects such as reliability and service punctuality (affected by infrastructure).
	Capac	Ability of the asset(s) to support certain operational requirements (volume, speed, size, load, etc.)
	Obsol	Process of assets becoming antiquated in comparison with newer versions
	Environ	Environmental externalities derived from the asset life cycle activities
	Condit	Level of asset integrity, affecting the probability of failure
	Conseq	Consequences derived from asset failure
Relevant stakeholders	Safety	Protection of passengers, workers, and other people from physical harm or security breaches
	Utilis	Demand for services provided by the assets
	Know 1	Ability to use information in decisions and actions, and the impacts on process efficiency and effectiveness
	HR	Development of workers’ capabilities, career, and job satisfaction
	Access	Capacity of users to access the service (accessibility) and the ability of infrastructure to connect with other assets/services/systems (interoperability)
	Image	Overall corporate image, communication, and transparency

¹ This criterion includes the impacts on knowledge and process efficiency. This was decided based on two reasons: simplification of the value framework; the two dimensions showed high correlations with the topic “asset knowledge enablers”, demonstrated in the report “Criteria for evaluating asset management indicators” [40].

3.3. Analysis Criteria

The criteria used to analyse the value of each opportunity are defined in this step. The criteria are aligned with the value framework and customised to the purpose and scope. Each criterion is described in terms of performance levels (

Table 8. Performance levels used to analyse the context of digital twinning opportunities.

Criterion (Context)	Description of Performance Levels					Relative Value
LoS		Rail		Road		Rail	Road
		S1 (high performance/commuter)		S1.1 (high performance)		100	100
		S2 (high performance)		S1.2 (high performance)		80	80
		S3 (medium performance)		S2.1 (medium performance)		50	50
		Various		S2.2 (medium performance)		50	50
		S4 (low performance)		Various		0	50
				S3 (low performance)			0
Subcl 1	Relevance	Rail		Road
	High	Rail track Turnout Bridges, tunnels, viaducts Operational control centre Track platform Interlocking and external equipment		Pavement Drainage Safety barriers Traffic lights Lighting Other equipment (kerbs, etc.)		100
	Medium	Level crossings Catenary ATP system Platform protection Substations and sectioning posts Other track equipment Elevated crossings Complementary safety systems Culverts and drainage Operation support systems Transmission Cable trays and transmission support		Viaducts, underpasses Roadside and pavement Bridges, tunnels Vertical signalling Culverts Platform protection Overpasses Fences Variable-message signs Emergency Communications System (SOS) Video surveillance		50
	Low	Passenger platform Buildings Current return paths and earthing system Energy remote control and supervisory systems Passenger information system Acoustic protection, fences		Acoustic protection Other built elements (car parks, rest areas, etc.) Traffic counting Gantries and tolls Buildings		0
Invest	System of Systems	Asset system	Asset	Asset component	Asset subcomponent
	]0;1 mEUR[	]0;500 kEUR[	]0;50 kEUR[	]0;10 kEUR[	]0;1 kEUR[	100
	[1 mEUR;5 mEUR[	[500 kEUR;1 mEUR[	[50 kEUR;100 kEUR[	[10 kEUR;50 kEUR[	[1 kEUR;5 kEUR[	50
	[5 mEUR;10 mEUR[	[1 mEUR;5 mEUR[	[100 kEUR;500 kEUR[	[50 kEUR;100 kEUR[	[5 kEUR;10 kEUR[	0
	[10 mEUR;50 mEUR[	[5 mEUR;10 mEUR[	[500 kEUR;1 mEUR[	[100 kEUR;500 kEUR[	[10 kEUR;50 kEUR[	−50
	[50 mEUR;+∞[	[10 mEUR;+∞[	[1 mEUR;+∞[	[500 kEUR;+∞[	[50 kEUR;+∞[	−100
Financ	European Union funding					100
	European Investment Bank financing					50
	Self-financing					0
Synerg	Leverages investments already implemented					100
	There are planned investments related to the opportunity					50
	There are no relevant synergies					0
Compl	Court summons/instruction from grantor/legal obligation					100
	Recommendation from legal advice/national or international commitments					66.7
	Municipal or local commitments					33.3
	No requirements or commitments					0

Note: ||—“neutral” level of performance; ||—“good” level of performance. ¹ Based on two inquiries performed in IP, between 9 and 23 March 2017 [41,42].

).

The user can choose the most adequate level to characterise the context in which the opportunity occurs and assess the expected impacts of each opportunity. The levels highlighted in light green correspond to intrinsically “good” (undoubtedly satisfying) levels of performance, and the light blue ones correspond to intrinsically “neutral” (neither satisfying nor unsatisfying) levels of performance. The use of “good” and “neutral”, as described, is known to contribute significantly to the intelligibility of each criterion [34].

Like risk analysis, the analysis of digital twinning opportunities may be influenced by the quality of information, divergence of opinions, biases, perceptions, and judgements [33]. Analysing the impacts of digital twinning opportunities requires an adequate trade-off between the level of detail required to analyse each impact and the inherent uncertainty. A rating scale is used to achieve this. Such a scale is aligned with the expected utility theory, in which the expected value (utility) combines the effect of probabilities (likelihood) and the consequences (impacts) [39]. A 5-category rating scale (

Table 9. Rating scale used to analyse the impacts on assets and stakeholders.

Criterion (Impacts)		Description of Performance Levels	Score	Relative Value
Asset(s)	CAPEX, OPEX, Avail, Capac, Obsol, Environ, Condit, Conseq
		Significant or highly likely positive impact	++	100
		Marginal or likely positive impact	+	50
		Negligible or unknown impact	+/−	0
Stakeholders	Safety, Utilis, Know, HR, Access, Image	Marginal or likely negative impact	−	−50
		Significant or highly likely negative impact	− −	−100

Note: ||—“neutral” level of performance; ||—“good” level of performance.

) was chosen to analyse the impacts according to subjective judgements, following two main objectives: have a number of scale points that address the previously discussed aspects without being too fine or too coarse; and have an odd number of categories, in which the middle category can represent neutral, negligible, or unknown impacts [35].

Digital twinning opportunities may refer to different asset hierarchical levels. As the value generated from the organisation occurs at the macro level (overall value), the value measurement at this level must be somehow inferred from the impacts measured at the applicable asset level (local value) [16]. Scale factors were used to bridge this gap. These factors (

Table 10. Scale factors used to obtain the overall value of opportunities.

Criterion (Context)	Asset Hierarchical Level	Scale Factor
Hierar 1	System of Systems	1.00
	Asset system	0.50
	Asset	0.25
	Asset component	0.15
	Asset subcomponent	0.10

¹ Aligned with the “hierarchy” dimension of the UNI-TWIN model (see Section 2.3).

) express higher or lower impacts of different asset hierarchical levels in the organisational and asset management objectives. They can be further tuned using an analysis with a group of different case studies. Choosing the most representative level for analysing each opportunity may depend on relevant aspects such as the asset register (asset breakdown structure), network redundancy (perhaps differentiating between rail and road networks), operational costs, and utilisation, among others.

Following the principles of the multi-criteria decision theory, the overall utility (value) of a given opportunity is a function of the weights of each value criterion [43].

As illustrated in

Table 11. Weights attributed to the analysis criteria.

Value Criteria/Weights									Total
Context	Hierar	LoS	Subcl	Invest	Financ	Synerg	Compl	Hierar
	- 1	3%	2%	2%	7%	2%	2%	- 1	18%
Asset(s)	CAPEX	OPEX	Avail	Capac	Obsol	Environ	Condit	Conseq
	12.5%	13%	12.5%	2%	2%	3%	5%	5%	55%
Relevant stakeholders	Safety	Utilis	Know	HR	Access	Image
	13%	5%	2%	1%	4%	2%			27%
Total									100%

¹ Not applicable to “Hierarchy” (see Table 10).

, an exception was established regarding the weight distribution. IP experts considered that if a given opportunity is part of a court summons, an instruction from the grantor, or a legal obligation, it should be directly considered for implementation. The weighting system is overridden in such cases (i.e., the weight of compliance automatically becomes “100%”). The impact of alternative weighting systems and analysis methods on the overall results can be further studied in the future, but this is out of the scope of this research.

4. Results and Discussion

4.1. Analysis of Context

Table 12. Relative and weighted value generated by the context of each opportunity (measured at the applicable level).

Case Study		Asset(s) Context						Opportunity Context			Context
		Network	Hierar	LoS	Subcl	Invest		Financ	Synerg	Compl
[A]	Score	Rail	Asset subcomp.	S1	Substations and sectioning posts (medium relevance)	9.5 kEUR	[5 kEUR; 10 kEUR[	European Union funding	There are no relevant synergies	No requirements or commitments
	Relative value		0.10 *	100	50		0	100	0	0
	Weighted value			3.0	1.0		0.0	7.0	0.0	0.0	11.0
[F]	Score	Rail	Asset system	S1	Various subclasses (high relevance)	2.5 mEUR	[1 mEUR; 5 mEUR[	European Union funding	There are planned investments related to the opportunity	No requirements or commitments
	Relative value		0.50 *	100	100		0	100	50	0
	Weighted value			3.0	2.0		0.0	7.0	1.0	0.0	13.0
[K]	Score	Rail and Road	Asset system 1	Various levels 2	Various subclasses (high relevance)	50 k€	]0;500 k€[	Self- financing	There are no relevant synergies	No requirements or commitments
	Relative value		0.50 *	50	100		100	0	0	0
	Weighted value			1.5	2.0		2.0	0.0	0.0	0.0	5.5
	Weights			3%	2%		2%	7%	2%	2%	18% (total)

¹ After consultation with experts, the author assumed that the group of 18 infrastructure assets was wide enough to represent an asset system and, consequently, to have a level 3 in terms of Hierarchy in the UNI-TWIN model and the value framework. ² The 18 infrastructure assets have various performance levels, from S4(rail)/S3(road) to S1 (rail)/S1.1 (road). Following some works [44] and internal guidelines [45], a partial value of 50 (equivalent to medium performance) was given to such scenarios. * See Table 10.

presents the results attained by applying the analysis criteria applicable to the context (see Table 8) of case studies [A], [F], and [K] of IP (see Section 2.2). Considering the scores and the criteria weights defined in Table 8, the context value of each opportunity is calculated at the applicable hierarchical level. Table 12 shows the results of these calculations at the applicable hierarchical level, i.e., without factoring the relative value (scale factor) according to the asset hierarchical level. The hierarchical level affects all dimensions of the analysis and the overall result (see Section 4.4), as discussed in Section 3.3.

While case study [A] focuses on the insulating oil of a power transformer (an asset subcomponent), case studies [F] and [K] respectively address the project of a new high-speed rail line and the satellite monitoring of infrastructure assets (asset systems). Case studies [A] and [F] refer to assets performing at the highest level of service within the rail network (S1). Case study [K] is a mixed case of performance levels (see footnote 4).

According to the public contract “Instalação de equipamento de análise de gases para transformadores de potência na subestação de Salreu” [46], the acquisition and installation of the transformer oil monitoring equipment (case study [A]) has a cost of EUR 9500. Case study [K] has an estimated up-front cost of EUR 50,000, including set-up and installation fees. Case study [F] has an estimated investment need of EUR 2.5 M, resulting from a 6-year implementation period with yearly costs (e.g., recruitment of BIM specialists, Autodesk Revit licenses, consulting services) and non-recurring costs (e.g., IT infrastructure, BIM training).

While opportunities [A] and [F] receive European financing (the best scenario in terms of “financing”—relative value “100”), [K] is currently dependent on self-financing (“0”). No opportunity serves any kind of legal requirements or political commitments, resulting in the minimum score in “compliance”. Case study [F] is integrated into a major investment project regarding the construction of a new high-speed rail line, and so, contrary to what happens with case studies [A] and [K], there are planned investments related to this opportunity (value of “50” in “synergies”). There were no requirements or commitments related to any of the opportunities.

As Table 12 shows, opportunity [F] shows the highest relative value in terms of context (13.0), followed by case studies [A] (11.0) and [K] (5.5). None of the case studies shows negative results in terms of context.

4.2. Impacts on Asset Cost, Risk, and Performance

The rating scale presented in Table 9 is used to analyse the impacts of each opportunity on the applicable asset hierarchical levels.

Table 13. List of impacts on cost, performance, and risk identified for each case study.

Case Study	Impact (Description)	Metrics (Examples)	Impact Criterion	Score
[A]	Reduction in laboratory testing periodically performed to the properties of the insulating oil	Reduction of 1 test every 2 years (2750 EUR/2 years): 2500 EUR/test (power transformer of type A/B with oil in acceptable condition) + 250 EUR/test for expert supervision	OPEX	++
	Reduction in oil replacements caused by the decrease in oil extractions during laboratory testing	Reduction of circa 3 L of insulating oil per sample, equivalent to EUR 90 per sample/test	OPEX	+
	Reduction in the environmental impact caused by the decrease in oil replacements	Reduction of circa 3 L of insulating oil per sample, equivalent to 6 kg CO2/sample/test [47]	Environ	++
	Reduction in fault risk by early detection of degradation patterns in the insulating oil	Note: The reduction in the need for expert supervision and contractor services (number of trips) and the impacts of the monitoring equipment could also be considered	Cond Conseq	+ +
	Increased data quality and confidence in decision-making	It promptly alerts the user to abnormal readings, minimising the probability of unplanned asset failures (increased costs of reactive interventions).	OPEX Environ	+ +
	Increase in recurring costs due to the operation and maintenance of the monitoring equipment	Reduction in the number of confirmatory tests needed to validate previous tests (often producing unreliable results caused by various error sources, e.g., sample extraction proceeding)	OPEX	−
[F]	Increased data quality and confidence in decision-making	A more precise estimate of additional work and a decrease in rework caused by planning and design phases with fewer errors contribute to reducing capital expenditure.	CAPEX	+
	Reduction in the environmental impact of renewals and maintenance actions	A collaborative and coordinated work methodology as BIM decreases the risk of errors and rework (e.g., fewer clashes, delays), contributing to the reduction of the environmental impact (e.g., material consumption, equipment use, waste)	Environ	+
	Reduction in downtime due to faults and maintenance work	Better informed interventions consume less time (limited and costly in rail operations—MTTR, for example) and produce fewer errors (less rework)	Conseq	+
	Increase in recurrent costs due to the adoption of BIM methodology	Need for periodic training of human resources, payment of IT licenses (e.g., Autodesk Revit), and operation (e.g., electricity consumption, data storage, data quality review) and maintenance of IT infrastructure	OPEX	−
[K]	Reduction in fault risk by early detection of increasing displacement patterns	Automatically alerts users once safety thresholds are crossed, minimising the probability of unplanned asset failures (increased costs of reactive interventions).	Cond Conseq	+ ++
	Optimisation of inspection activities	The use of a complementary monitoring system helps asset managers prioritising inspections in a more informed way, minimising time and resource consumption	OPEX	+
	Increase in recurring costs due to the operation of the monitoring system (service provider)	Costs associated with the operation of the monitoring system provided by an external entity.	OPEX	−

presents a non-exhaustive list of impacts on asset cost, performance, and risk identified by IP experts for the case studies [A], [F], and [K].

Based on the impacts listed in Table 13, the analysis results are presented in

Table 14. Relative and weighted value generated by impacts on asset(s) (measured at the applicable level).

Case Study		Costs				Performance		Risk		Asset(s)
		CAPEX	OPEX	Avail	Capac	Obsol	Environ	Condit	Conseq
[A]	Score	+/−	++	+/−	+/−	+/−	++	+	+
	Relative value	0	100	0	0	0	100	50	50
	Weighted value	0.0	13.0	0.0	0.0	0.0	3.0	2.5	2.5	21.0
[F]	Score	+	−	+/−	+/−	+/−	+	+/−	+
	Relative value	50	-50	0	0	0	50	0	50
	Weighted value	6.3	-6.5	0.0	0.0	0.0	1.5	0.0	2.5	3.8
[K]	Score	+/−	+/−	+/−	+/−	+/−	+/−	+	++
	Relative value	0	0	0	0	0	0	50	100
	Weighted value	0.0	0.0	0.0	0.0	0.0	0.0	2.5	5.0	7.5
	Weights	12.5%	13%	12.5%	2%	2%	3%	5%	5%	55% (total)

. Considering the scores and the criteria weights (Table 11), the relative value for the impacts on assets of each opportunity can be calculated at the applicable hierarchical level (Table 10).

As Table 14 illustrates, case study [A] is the opportunity with the highest value (21.0) in terms of impacts on asset(s), which is the analysis dimension with the highest weight of the three (weight of 55%). Case studies [K] (7.5) and [F] (3.8) come in second and third places, respectively. One relevant factor that helps explain this difference is that case study [A] has a significant or highly likely positive impact (relative value “100”) on OPEX, which is the criterion with the highest weight (13%) from all analysis criteria (see Table 11). Moreover, case study [A] performs better or equally well as the others in almost all criteria (except for “Consequence”). Case study [F] is the best-performing opportunity in terms of CAPEX (weight of 12.5%) but is the only one with negative impacts on OPEX, which heavily penalises its weighted value, especially when compared to case study [K].

4.3. Impacts on Stakeholders

Table 15. List of impacts on stakeholders identified for each case study.

Case Study	Impact (Description)	Metrics (Examples)	Impact Criterion	Score
[A]	Increased data quality and confidence in decision-making	Reduction in the number of confirmatory tests needed to validate previous tests (often producing unreliable results caused by various error sources, e.g., sample extraction proceeding)	Know	++
	Increased safety caused by the reduction in in-person tests	Reducing the number of incidents and accidents caused by the presence of workers (service providers) performing the laboratory tests on the insulating oil	Safety	+
	Increased risk of data breaches	Increased exposure to data breaches and unauthorised access to asset data	Safety	−
	Increased risk of dependence on certain technology solutions and vendors	Adopting a work methodology for such a broad scope might create dependence on certain technologies or providers. Poor after-sales service, discontinued versions, increases in license price, and incompatibility between versions are issues that might arise.	Know	−
[F]	Increased data quality and confidence in decision-making	A more precise estimation of additional work and a decrease in rework caused by planning and design phases with fewer errors contribute to better knowledge.	Know	++
	Reduction in risk of losing data and information	Using digital infrastructure (e.g., cloud) and decentralised access to store and access data decreases the risk of permanent loss of documented information (e.g., design documents, as-is) by physical destruction (e.g., biological agents, fire, floods)	Know	+
	Safer work procedures due to better visualisation	During design, construction, maintenance, and disposal phases, work planning can benefit from immersive and comprehensive visualisation of the asset and its environment, producing safer procedures and a safer work environment (e.g., fewer accidents or less severe accidents)	Safety	+
	Increased risk of data security	Decentralised and shared access to data increases exposure to data breaches and unauthorised access to data related to critical assets	Safety	−
	Increased risk of dependence on certain technology solutions and vendors	Adopting a work methodology for such a broad scope might create dependence on certain technologies or providers. Poor after-sales service, discontinued versions, increases in license price, and incompatibility between versions are issues that might arise.	Know	−
	Increased risk of interoperability issues	When exchanging and updating file versions, interoperability issues might occur, leading to data losses and inconsistencies	Know	−
	Lack of some standards regarding the use of BIM in the rail and road context	The sector’s overall maturity regarding the use of BIM is currently low. A lack of standards regarding information requirements and data exchange protocols persists within the community. Due to the learning curve, the initial stages of development are expected to be less productive and more extended in time due to the learning curve.	Know	−
	Improved corporate image by adopting innovative work methodologies involving the use of technologies	The corporate image of IP may benefit from the adoption of BIM in a project of such importance as the new high-speed rail line (e.g., innovation awards, news in the media)	Image	+
[K]	Improvement of organisational knowledge	The satellite-based monitoring system complements other inspection activities by adding more data (e.g., longer time series).	Know	+
	Increased safety caused by the reduction in in-person surveys	Reducing the number of incidents and accidents caused by the presence of workers performing topographical surveys	Safety	+
	Improved corporate image by adopting innovative monitoring techniques	The corporate image of IP may benefit from the adoption of innovative monitoring techniques (e.g., news in the media)	Image	+

presents a non-exhaustive list of impacts on stakeholders identified by IP experts for the case studies.

Based on the impacts listed in Table 15, the results in terms of impacts on stakeholders are presented in

Table 16. Relative and weighted value generated by impacts on stakeholders (measured at the applicable level).

Case Study		Safety	Utilis	Know	HR	Access	Image	Stakeholders
[A]	Score	+/−	+/−	++	+/−	+/−	+/−
	Relative value	0	0	100	0	0	0
	Weighted value	0.0	0.0	2.0	0.0	0.0	0.0	2.0
[F]	Score	+/−	+/−	+	+/−	+/−	+
	Relative value	0	0	50	0	0	50
	Weighted value	0.0	0.0	1.0	0.0	0.0	1.0	2.0
[K]	Score	+	+/−	+	+/−	+/−	+
	Relative value	50	0	50	0	0	50
	Weighted value	6.5	0.0	1.0	0.0	0.0	1.0	8.5
	Weights	13%	5%	2%	1%	4%	2%	27% (total)

. Regarding the previous scores and the criteria weights (Table 11), the relative value for the impacts on stakeholders of each opportunity was calculated and measured at the applicable hierarchical level (Table 10).

In contrast to the previous results for the asset(s) scope, case study [K] shows a higher value (8.5) than case studies [A] and [F] (both with 2.0) regarding impacts on stakeholders at the local level. A better performance on “Safety and security” (6.5 against 0.0)—the criterion with the highest weight in this specific analysis (13%)—explains this result. No relevant impacts were identified in any case study regarding “Utilisation and revenues”, “Human resource development”, and “Accessibility and interoperability”. On the other hand, all case studies show positive impacts in terms of “Knowledge and efficiency”, as shown in Table 15 and Table 16. However, these impacts have little contribution to the overall value of each opportunity, since “Knowledge and efficiency” belongs to a group of criteria with the second lowest weight of all (2%), only behind “Human resource development” (1%) (Table 11). Future developments should include a sensitivity analysis of these variables.

4.4. Overall Value Analysis

Table 17. Summary of local and overall value analyses.

Case Study	Relative Weighted Value (without Scale Factor)				Scale Factor (Hierar)	Overall Weighted Value (with Scale Factor)
	Context	Asset(s)	Stakeholders	Result		Context	Asset(s)	Stakeholders	Result
					0.10
[A]	11.0	21.0	2.0	34.0		1.1	2.1	0.2	3.4
					0.50
[F]	13.0	3.8	2.0	18.8		6.5	1.9	1.0	9.4
					0.50
[K]	5.5	7.5	8.5	21.5		2.8	3.8	4.3	10.8
Interval	[−2; 18]	[−55; 55]	[−27; 27]	[−84; 100]		[−2; 18]	[−55; 55]	[−27; 27]	[−84; 100]

shows the combined result of the analyses of the context, impacts on asset(s), and impacts on stakeholders, both with and without considering the scale factor that expresses the hierarchical level applicable to the opportunity.

Despite the Real-time Dissolved Gas Analysis system (case study [A]) scoring relatively better than the BIM methodology for the new railway line (case study [F]) and the Displacement monitoring system using satellite data (case study [K]) at the applicable hierarchical level (34.0 against 18.8 and 21.5, respectively), [A] has the lowest overall result because its hierarchical level is significantly lower than [F] and [K] (asset subcomponent versus asset systems). The application of the corresponding scale factors makes case studies [K] and [F] stand out as the overall best options compared to case study [A] (10.8 and 9.4 against 3.4, respectively). A more detailed analysis reveals that the application of scale factors maintained the results for the best-performing cases in the analysis of context (case study [F]) and stakeholder impacts (case study [K]). However, in terms of impacts on assets, the top-performing case shifted from case study [A] to case study [K]. No case study showed negative results (value destruction) in any dimension of analysis (context; impacts on asset cost, risk, and performance; impacts on stakeholders), neither in relative weighted values nor in overall weighted values. This suggests that, according to the methodology used, each case study could add value to the organisation and its stakeholders.

These results should be interpreted considering the following notes:

• More important than the numerical results obtained, the use of three different case studies enabled the validation of the analysis methodology, which was the underlying aim of the study described in this chapter.
• Despite the adoption of BIM methodology for the new railway line (case study [F]) and the Displacement monitoring system using satellite data (case study [K]) showing a higher overall value than the Real-time Dissolved Gas Analysis system (case study [A]), it does not mean that they are necessarily “good” or viable digital twinning opportunities. This means that, according to this methodology and existing knowledge, case studies [K] and [F] are more likely to generate higher value for IP than [A].
• The analysis is performed at an intermediate level in terms of information detail to integrate and communicate the impacts of different opportunities related to different hierarchical levels. Some criteria could be further developed in terms of performance detail, such as OPEX (e.g., 10% reduction in asset OPEX), but these improvements should be aligned with the existing capacity to report such impacts.
• When applied to a larger set of opportunities, this analysis method is expected to assist asset managers in shortening the list of competing investments according to a given “cut-off line”.
• The proposed analysis method can be combined with other decision-support methods, tools, and metrics, such as Value/CAPEX, Life Cycle Costing, Net Present Value, Risk-based analysis, etc. [48]. Asset managers can select and assess the applicability and usefulness of such tools as those listed in [19,20,39].
• The analysis results are directly influenced by different sources of uncertainty, namely, judgements, impacts, scoring, scale factors, and weighting [49]. Future studies should consider these uncertainties further and their impact on the results. Using more case studies or comparisons with other tools (e.g., MACBETH) could help with this task.

The results obtained should be used to inform the evaluation phase, which involves comparing the results of the analysis with pre-established criteria to determine the actions required (do nothing, implement, undertake further analysis, reconsider, etc.) [33]. When applied to a larger set of opportunities, this analysis method is expected to assist asset managers in shortening the list of competing investments according to a given “cut-off line”. The evaluation stage will be the focus of future works.

5. Conclusions

This article discusses the need for a clear understanding regarding the value of digital twinning opportunities. The proposed method for analysing the value derived from digital twinning opportunities relies on the conceptual foundations of value laid down by ISO 55000. Therefore, its structure is conceptually prepared for application to other types of opportunities and organisational contexts with the necessary adjustments. The proposed methodology analyses opportunities and their different impacts on several criteria based on qualitative value judgements. It accommodates the uncertainty associated with those judgements and the multiple and conflicting objectives that typically exist in asset management (costs, benefits, and risks) through a comprehensive and aligned value framework. This method was applied to three rail and road infrastructure case studies identified by experts from Infraestruturas de Portugal, S.A.—a public body responsible for developing, operating, and maintaining road and railway networks in Portugal.

The use of three different case studies enabled the validation of the analysis methodology, which was the underlying aim of the study described in this chapter. This analysis method provided a structure of thought and a common language to facilitate the communication of digital twinning opportunities in terms of their different contexts and impacts. Moreover, this work contributed with a novel approach to analysing the value of digital twinning opportunities. Infrastructure asset management organisations and IP, in particular, are now equipped with a value-based analysis tool that can support investment decision-making and help derive the highest value from the asset portfolio. This value-based approach is a significant contribution considering the lack of an appropriate framework to support asset managers in planning the digital twinning journey, as highlighted in multiple studies related to DTs, such as [7,8,9,10,11].

Following the principles of the multi-criteria decision theory, the overall value of a given opportunity was calculated as a function of the weights of each value criterion. The results are influenced by different sources of uncertainty, namely, judgements, impacts, scoring, and weighting. Future studies should consider these uncertainties further and their impact on the results. Using more case studies or comparisons with other tools (e.g., MACBETH) could help with this task.

When applied to a larger set of opportunities, this analysis method is expected to assist asset managers in shortening the list of competing investments, considering the tight competition for investment and the need to extract the best value from the limited available resources. The proposed analysis method can be combined with other decision-support methods and tools, such as Life Cycle Costing, Net Present Value, and Risk-based analysis. Moreover, as this methodology is an example of an analytical approach, its results should be complemented with an intuitive analysis whenever possible.

The analysis of digital twinning opportunities may be influenced by the quality of information, divergence of opinions, biases, perceptions, and judgements. The results are case-specific as they depend on the aspects of the assets covered and the opportunities identified. However, the conclusions extracted from the proposed methodology and the assumptions made can be extended to any case study.