Abstract
Sustainability education increasingly requires learning environments that connect systems understanding with real-world decision-making. This study examines the integration of a district-heating digital-twin-based modelling environment into a multilevel living-lab process at Riga Technical University, Latvia. An exploratory mixed-method case study was conducted over 13 months and involved 24 professionals in the co-creation stage, 196 undergraduate students working in 98 pairs, 30 master's students working in 15 pairs, and eight external stakeholders. Data included World Café outputs, 294 undergraduate scenario responses, modelling-environment simulation outputs, master's student multi-criteria decision analyses, and stakeholder evaluations. Quantitative scenario outputs were examined in relation to emissions, costs, and system performance, while participant-produced materials were analysed to identify patterns in systems reasoning, trade-off recognition, and decision priorities. Undergraduate students used the modelling environment to explore relationships between technical parameters and environmental and economic outcomes. At the master's level, 12 of the 15 groups (80%) prioritised environmental performance when selecting their preferred scenarios, whereas external stakeholders assigned greater priority to investment cost and practical feasibility. This divergence suggests that the modelling environment may help make competing decision priorities visible and support structured discussion between learner and practitioner groups. The study is limited by its single-institution design, absence of a control group and pre-test/post-test learning assessment, limited stakeholder sample, and reliance on participant-produced work. The findings therefore provide preliminary case-based evidence of the educational and decision-support potential of digital-twin-supported living labs rather than evidence of general effectiveness.
1 Introduction
Addressing complex challenges such as climate change, energy-system transition, and urban decarbonisation requires technological innovation to be combined with new approaches to stakeholder participation, institutional collaboration, and knowledge creation (). Sustainability transitions are inherently socio-technical and require interdisciplinary thinking, systems understanding, and the ability to analyse interactions among environmental, economic, technological, and institutional factors (). Higher education institutions can contribute to these transitions by preparing students to work with complex systems and by connecting education, research, and societal engagement. However, conventional pedagogical approaches may provide limited opportunities for students to engage with real-world sustainability challenges and the competing interests involved in their resolution (; ). Digital, experiential, and collaborative learning environments may help address this limitation by enabling students to explore system relationships, test alternative solutions, and connect academic knowledge with practical decision contexts.
Living labs provide context-rich environments in which students, researchers, university staff, public authorities, companies, and other societal actors can collaborate on real-world sustainability challenges. Their commonly recognised characteristics include experimentation in real-life settings, active user involvement, co-creation, multi-stakeholder participation, and iterative learning processes (; ). In higher education, campus living labs can connect teaching and research with campus operations and external engagement. conceptualise campus living labs as systems for innovation and learning in which experimental and research activities are co-produced by students, teachers, researchers, and operational staff. Depending on their purpose and institutional organisation, such initiatives may perform educational, test-bed, strategic, or grassroots functions. Applied examples include campus-based experimentation with energy-efficiency measures, renewable-energy systems, resource management, and organisational innovation (; ; ).
The educational value of living labs is associated with experiential, project-based, and challenge-based learning. These approaches provide opportunities for students to work on authentic problems, interact with actors outside the classroom, and develop interdisciplinary and collaborative capacities. Living-lab activities have been used to introduce university students to co-creation and responsible research and innovation through experiential learning (). Challenge-based sustainability education can similarly help students connect theoretical knowledge with contextualised environmental and circular-economy problems (). In the energy-transition field, networked learning has been applied to prepare future professionals for interdisciplinary and practice-oriented work (), while project-based approaches have been implemented in postgraduate sustainability planning and renewable-energy design education (; ).
Despite this potential, living labs are not always fully integrated into the core structures of higher education. A systematic review of 93 university-governed living-lab experiences found that many developed as bottom-up initiatives at the institutional periphery and required stronger governance, stakeholder relationships, capacity building, and interdisciplinary coordination to become embedded in university activities (). Reviews of living labs in sustainability transitions likewise identify continuity, institutional support, stakeholder participation, and systematic evaluation as important enabling conditions (). Their contribution also depends on the collective capabilities that participants develop and on the institutional and territorial contexts in which the initiatives operate (). More recent evidence from 18 international university living labs indicates that their scope and longevity are often constrained by temporary project logic and established academic and operational structures, although some initiatives overcome these barriers through stronger university-wide institutionalisation ().
From an educational perspective, the available evidence remains fragmented and is frequently based on individual courses or singular case studies. show that sustainability-oriented living labs can provide authentic, collaborative, and context-rich learning environments, but that the presence of a real-world setting does not automatically lead to sustainability-oriented learning. Outcomes also depend on pedagogical design, facilitation, reflection, stakeholder interaction, and the mechanisms through which students interpret their experiences. Consequently, less is known about how living-lab processes can be integrated across different educational levels and implemented with larger student cohorts while maintaining meaningful interaction with practitioners and external stakeholders. This constitutes the first gap addressed in the present study.
Digital-twin technologies offer an additional means of connecting real-world systems, analytical models, and educational activities. A digital twin may be broadly described as a virtual representation of a physical asset, process, or system that combines relevant data, computational models, and analytical functions (; ). However, digital twins differ substantially in maturity and functionality. distinguish digital-twin implementations according to dimensions such as context, data, computing capabilities, model functions, system integration, control, and human–machine interaction. In energy systems and Positive Energy Districts, digital twins and related modelling environments can support scenario analysis, resource-efficiency assessment, system optimisation, and the examination of alternative technical configurations (; ).
Digital twins have also been discussed as tools for engineering and experiential education. Conceptual work has highlighted their potential to connect digital representations with system analysis and practical learning activities (; ). Course-based applications have introduced students to digital-twin technology through engineering tasks and have reported potential benefits related to motivation, understanding, and the connection between theory and practice (). Digital environments may also provide relatively controlled settings in which users can test assumptions and explore technical and socio-economic alternatives without intervening directly in the physical system ().
Nevertheless, the empirical evidence concerning digital twins as educational tools remains heterogeneous. A systematic review of digital-twin and virtual-learning environments in smart manufacturing education identified potential benefits but also continuing limitations related to implementation barriers, comparability of assessment methods, and evidence concerning long-term learning outcomes (). A review of digital twins for education, training, and learning in the construction sector similarly found a fragmented and domain-specific evidence base (). Some empirical studies have evaluated digital-twin applications using usability measures, interviews, task performance, workload assessment, knowledge interpretation, or comparative training designs. For example, used a mixed-method approach to evaluate the usability and acceptance of a patient digital twin in critical-care education, while compared traditional and digital-twin-supported robotic training using measures related to task performance, interpretation, user experience, and knowledge retention. However, much of this evidence remains concentrated in specialised technical or professional-training contexts. Broader evidence concerning how digital twins support systems understanding, sustainability trade-off analysis, and structured decision-making across multiple levels of higher education remains limited. This represents the second gap addressed in this study.
Combining living-lab processes with digital modelling creates a socio-technical setting in which stakeholder knowledge, educational activities, and scenario analysis can be connected. The living-lab component provides an authentic problem context and opportunities for co-creation, whereas the digital model provides a controlled environment for testing assumptions and examining alternative system configurations (; ). Such integration may enable feedback between practitioner knowledge, model development, educational use, and stakeholder assessment. However, its value depends on how participant contributions are collected, how they inform the model or educational tasks, and how the resulting outputs are evaluated. It should therefore not be assumed that combining the two approaches automatically produces improved learning or better decisions.
A further issue concerns the extent to which learners and professional stakeholders evaluate sustainability problems in similar ways. Sustainability-related decisions require participants to balance environmental performance, economic cost, technical feasibility, reliability, and implementation constraints. These dimensions may be prioritised differently because learners and practitioners operate within different institutional roles, levels of experience, responsibilities, and time horizons. , comparing 204 students with 172 professionals, found that student evaluations did not strongly predict what professionals valued in sustainable-design practices, although similarities were observed in the reasons given for their judgements. More recent multi-stakeholder research also indicates that students, teaching staff, and employers may assign different importance to sustainability-related competences and decision dimensions (). Differences between academic and practitioner perspectives have likewise been observed in sustainable-project selection, particularly regarding the importance of cost, experience, project maturity, uncertainty, and technology transfer (). However, few studies have compared how learners and external stakeholders evaluate the same digitally modelled sustainability problem. This constitutes the third research gap.
On the basis of the reviewed literature, three connected gaps are therefore identified. First, although living labs are increasingly used in higher education, less is known about their integration across educational levels and their implementation with larger student cohorts while maintaining meaningful practitioner and stakeholder involvement. Second, educational applications of digital twins are growing, but the empirical evidence remains concentrated in specialised technical contexts and provides limited insight into systems understanding and sustainability-related decision-making. Third, few studies compare how learner and external stakeholder groups prioritise environmental, economic, and technical criteria when evaluating the same digitally represented sustainability challenge.
This study examines how a district-heating digital-twin-based modelling environment was integrated into a multilevel living-lab process at Riga Technical University. The 13-month case involved practitioner co-creation, undergraduate scenario exploration, master's-level multi-criteria evaluation, and external stakeholder assessment. The aim was to explore how this combined approach can support the interpretation of system relationships, analysis of sustainability trade-offs, and comparison of decision priorities among learner and practitioner groups.
The study makes three case-based contributions. First, it documents a four-stage process for integrating practitioner co-creation, undergraduate education, master's-level analysis, and external stakeholder evaluation around one district-heating digital model. Second, it examines participant outputs for evidence of system interpretation and recognition of environmental, economic, and technical trade-offs. Third, it compares the decision priorities of master's student groups and external stakeholders when evaluating district-heating transition scenarios. Figure 1 presents the analytical relationship between the living-lab process, the digital model, the educational activities, and the comparison of participant decision priorities.
Figure 1
The study addresses the following research questions:
RQ1. How was the district-heating digital-twin-based modelling environment integrated into a multilevel living-lab process involving practitioners, undergraduate students, master's students, and external stakeholders?
RQ2. What evidence of system understanding and recognition of environmental, economic, and technical trade-offs is present in undergraduate and master's student outputs?
RQ3. How do the decision priorities of master's student groups differ from those of external stakeholders when evaluating district-heating transition scenarios?
2 Methods
This mixed-method case study investigates the integration of living lab approaches and digital twin technologies into educational and decision-making practices related to sustainability (; ). This research is grounded in a socio-technical systems perspective, recognising that sustainability challenges emerge from the interaction between technological infrastructure, human actors, and institutional frameworks (). This mixed-method case study combined modelling-environment outputs with participant-generated materials and stakeholder input to examine the implementation of the four-stage living-lab process, participants' interpretation of system relationships and sustainability trade-offs, and differences in decision priorities between learner and stakeholder groups.
In this study, the term digital-twin-based modelling environment refers to a scenario-based digital representation of the district-heating system used for simulation, analysis, and participant interaction. The system did not include continuous real-time data exchange with the physical system and should therefore not be interpreted as a fully operational digital twin.
The mixed-method design combined quantitative and qualitative evidence to address different aspects of the three research questions. Quantitative evidence comprised model simulation outputs, including emissions, costs and system-performance indicators, together with descriptive counts and scenario preferences across participant groups. Qualitative evidence comprised World Café materials, undergraduate written scenario responses, master's reports and stakeholder evaluation materials, which were reviewed for explicit references to system relationships, sustainability trade-offs and decision priorities. The two forms of evidence were complementary: simulation outputs characterised the consequences of alternative system configurations, whereas participant-generated materials provided insight into how these outputs were interpreted and used in educational and decision-making tasks. The findings were integrated descriptively rather than through statistical mixed-method inference.
The living lab was developed as an integrated multilevel setup for learning, modelling, and stakeholder involvement. This integrated system of living lab processes and digital modelling is hereafter referred to as the platform. The living lab was designed to support the staged development and use of a district-heating digital model across four participant groups. The analysis focused on how the model was incorporated into co-creation and educational activities, what evidence of system interpretation and trade-off recognition appeared in participant outputs, and how decision priorities differed between master's student groups and external stakeholders. This study combines the principles of co-creation with large-scale educational implementation and stakeholder assessment to examine participant-generated evidence of system interpretation, sustainability trade-offs, and decision-making alongside system-performance outputs (). The different levels allow examination of how participants engage with system relationships, analytical tasks, and practical decision-making across educational and professional contexts.
Undergraduate students participated in the educational application of the developed digital-twin-based modelling environment through scenario-based activities, while master's students and external stakeholders were involved in the analytical and decision-support stages of the study.
2.1 Co-creation process using the world café method
The district-heating digital model was developed in collaboration with professionals from municipalities and energy companies. The model was discussed and refined using the World Café method for structured cooperation and was used to study aspects of system design and the sustainability transition of district heating systems through dialogue (; ). The World Café method was used to elicit knowledge and incorporate stakeholders' viewpoints into the modelling process.
The session was held at eight thematic discussion tables and lasted roughly 2 h. The participants rotated between the tables for three discussion rounds. During the rounds, they refined and deepened their thoughts by discussing them with others at the same table. The tables address questions regarding the overall structure of the model and its functions, including which system components should be represented, which input and output parameters are relevant to decision-making, and which technologies and development pathways should be considered.
Throughout the session, the participants recorded ideas in the form of notes and posters regarding the system configuration, variables, and scenario concepts. The output from all the tables was compiled, and four system concepts were discussed to identify key priorities. These four system concepts were then used to define the model boundaries, variables, and performance indicators for the scenario pathways. The principal model outputs used in the subsequent analyses were emissions, costs, and system-performance measures. These indicators were generated by the model simulation for each scenario configuration and were then interpreted by the participating student groups and used in the later multi-criteria and stakeholder evaluations.
The digital model was developed with practitioner input to incorporate information considered relevant to the district-heating context. Potential variables affecting the future management of buildings and dwellings were identified and incorporated into the model. These include demand variability, integration of renewable energy, potential for electrification, and thermal storage. Stakeholder-defined constraints and priorities were used to develop the scenarios and evaluation criteria. Practitioner input was used to inform the model structure, scenario parameters, and evaluation criteria. At this stage, no formal usability or user-needs assessment was conducted; the contribution of the practitioners was limited to informing the content and practical context of the model.
2.2 Staged living lab implementation
The living lab was set up and conducted in a four-stage process, with different user groups in each stage and an increasing level of involvement (). Using this process, the digital model can be developed and applied step-by-step in real-life contexts, including educational settings. First, the initial model structure and assumptions were co-created with several practitioners. Subsequently, the developed model was used in a large-scale exploratory educational activity with 196 undergraduate students working in 98 pairs, allowing students to examine energy-system relationships and potential pathways towards improved energy efficiency and climate-neutrality objectives.
After the initial application of the model, the 30 master's students in environmental engineering working in 15 pairs were tasked with assessing and evaluating the model from a decision-making perspective, this time in greater depth and with more complex methods, in what was designated as stage three.
External stakeholders were consulted to evaluate whether the proposed model was relevant and useful for actual planning. The model was developed through a staged, feedback-informed process. In Stage 1, practitioner input was used to define the model boundaries, variables, technological options, and scenario structure. In Stage 2, the model was applied in the undergraduate scenario activity, providing evidence on how the model outputs and system relationships were interpreted. In Stage 3, master's students used the same modelling framework for structured multi-criteria comparison of alternative scenarios. In Stage 4, external stakeholders assessed the scenarios and model outputs in relation to practical relevance, cost, feasibility, reliability, and implementation considerations. Feedback from the successive stages informed the interpretation and use of the model in the following stages. The study did not systematically record a fixed number of model-development iterations or individual design pivots; therefore, the process is described as staged and feedback-informed rather than as a formally documented iterative cycle.
The four-stage implementation process of the living lab approach, including the participant groups, activities, and outputs, is illustrated in Figure 2.
Figure 2
2.3 Task design and interaction with the digital modelling environment
The learning activities were based on a scenario-based task in which the participants worked through a simulation experience utilising a digital modelling environment of the district-heating system. They altered system parameters to test their impacts on the system and completed a task based on a real-world scenario in which emissions needed to be reduced within an economic budget, while also requiring reliable operation. This scenario-based approach allows participants to explore the trade-offs between environmental, economic, and technical factors in a controlled yet realistic simulation environment.
The task was conducted at the undergraduate level with 196 students working in 98 pairs. Each pair was asked to develop three scenario responses. The task was designed as an exploratory learning activity within an educational simulation environment. There was no intention to fine-tune parameters to identify a single ‘optimal’ solution; rather, students explored different system configurations and considered the resulting system dynamics and interdependencies.
Several parameters, such as connection to district heating, integration of heat pumps or solar collectors, size of the thermal storage tanks, and operational parameters (temperature regimes and tariffs), could be varied by the students. Subsequently, the resulting emissions, costs, and system performance could be examined.
The participants, working in pairs, provided responses to questions regarding scenario selection, benefits, trade-offs, and final recommendations for addressing scenario challenges. These responses provided evidence of how the student pairs interpreted system behaviour and approached the decision-making task. In contrast to the structured multi-criteria analysis used at the master's level, the focus for undergraduate students was on interpreting the output from pre-run models to gain insight into how digital tools can facilitate sustainability-related decision-making. In this context, a digital modelling environment is employed as a pedagogical tool to support experiential learning and facilitate systems understanding for non-specialist end users.
For the master's-level activity, 30 students working in 15 pairs used the same modelling framework to develop and compare three district-heating scenarios: an environmentally oriented scenario, a cost-oriented scenario, and a balanced scenario. Each pair evaluated the alternatives using the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) across the environmental, economic, and technical indicators included in the model (). In TOPSIS, alternatives are ranked according to their relative distance from a positive ideal solution and a negative ideal solution. In this study, the positive and negative ideal solutions were derived from the best and worst values, respectively, for each indicator among the scenarios evaluated by each student pair; they were not independently assigned by the students or external stakeholders. Equal weights were assigned to all indicators. After the TOPSIS ranking of the three alternatives, each pair selected and justified its final preferred scenario, thereby addressing the trade-offs among environmental, economic, and technical objectives. This shift in task design enabled the modelling environment to function as both an exploratory tool and a platform for structured decision-making.
2.4 Data collection and analysis
This study combined model outputs with participant-generated materials and stakeholder input. The data sources included 294 undergraduate scenario responses, model simulation outputs, master's scenario reports and multi-criteria evaluations, World Café notes and posters, and evaluation material from the external stakeholder stage.
A total of 196 undergraduate students worked in 98 pairs. Each pair was asked to develop three scenario responses. In these written responses, students described their preferred system configuration, identified benefits and trade-offs, and provided a recommendation for addressing the assigned scenario challenge. The undergraduate analysis focused on evidence of how participants interpreted model outputs, identified relationships between system parameters and performance indicators, and recognised environmental and economic trade-offs. Model outputs were examined descriptively in relation to emissions, costs, and system performance.
The qualitative material comprised notes and posters produced during the World Café session, undergraduate written responses to the scenario tasks, master's scenario reports and justifications, and evaluation material from the external stakeholder stage. These materials were reviewed descriptively. The analysis focused on explicit references to system relationships, environmental and economic trade-offs, interpretation of model outputs, and decision priorities. Undergraduate responses were examined for references to emissions, costs, technical parameters, system interactions, and trade-offs between performance criteria. For the undergraduate analysis, each of the 294 scenario responses was reviewed for the presence of five non-exclusive descriptive categories: environmental effects, economic effects, technical or system relationships, recognition of trade-offs, and integrated recommendations. The unit of analysis was the individual scenario response rather than the individual student. The 294 responses were classified by two members of the research team using the predefined descriptive categories. Where a response was ambiguous, it was discussed within the research team and assigned only when the relevant category was explicitly supported by the response content. Environmental effects included explicit references to emissions, renewable energy, or climate-related outcomes; economic effects included costs, tariffs, investment, or operating expenses; technical or system relationships referred to explicit links between changes in system parameters and resulting system behaviour; trade-off recognition referred to acknowledgement of competing environmental, economic, or technical outcomes; and integrated recommendations included recommendations justified using at least two of these dimensions. A response could be assigned to more than one category. Frequencies and percentages were calculated descriptively for each category.
At the master's level, 30 students worked in 15 pairs. Each pair developed and evaluated three district-heating scenarios: an environmentally oriented scenario, a cost-oriented scenario, and a balanced scenario. TOPSIS was used to compare the alternative scenarios according to environmental, economic, and technical indicators. Equal weights were assigned to all indicators so that the rankings reflected scenario performance across the complete set of indicators rather than predetermined differences in criterion importance. Following the TOPSIS ranking of these three alternatives, each pair selected and justified a final preferred scenario. The preferred scenarios and the reasoning provided in the reports were used in the subsequent descriptive comparison of master's student decision priorities.
For RQ3, decision priority was operationalised from the final preferred scenario and the accompanying written justification rather than from the TOPSIS indicator weights. A group was classified as environmentally oriented when its written justification explicitly identified emissions reduction or broader environmental performance as a principal reason for selecting the preferred scenario. In these cases, the selected scenario also showed the strongest environmental performance among the alternatives considered, for example through lower emissions and/or a higher renewable-energy share. The classification therefore reflects scenario preference and the reasoning provided by the group, not differential weighting of TOPSIS criteria.
Eight external stakeholders participated in the final evaluation stage: two representatives from national ministries, two representatives from municipalities, and four representatives from energy companies. Their priorities were identified from discussion notes produced during the stakeholder evaluation. The notes were reviewed descriptively for explicit references to environmental performance, investment cost, technical feasibility, reliability, and practical implementation considerations. Their evaluations were used to identify the criteria emphasised by the participating stakeholder roles and to compare these decision priorities descriptively with those of the master's student groups.
The quantitative and qualitative evidence was integrated descriptively. Simulation outputs characterised the consequences of alternative system configurations, while participant-generated materials provided evidence of how these outputs were interpreted and used in educational and decision-making tasks. The analysis was exploratory and descriptive; no inferential statistical tests were performed.
2.5 Framework for linking research questions, data sources and analysis
An analytical framework, operationalised through the analysis methods summarised in Table 1, was developed to organise the evidence generated during the four-stage living-lab process and to link the participant groups, data sources, analytical procedures, and research questions. For RQ1, evidence from the World Café, model-development process, educational implementation, and stakeholder stage was used to describe how the digital modelling environment was integrated across the four stages of the living lab. For RQ2, undergraduate written responses, model simulation outputs, and master's scenario reports were examined for evidence of system interpretation and recognition of environmental, economic, and technical trade-offs. For RQ3, the preferred scenarios and decision priorities of the master's student groups were compared descriptively with those identified in the external stakeholder evaluations.
Table 1
| Research question | Participant group | Data sources and instruments | Analysis method | Main output |
|---|---|---|---|---|
| RQ1. | 24 professionals; 196 undergraduate students in 98 pairs; 30 master's students in 15 pairs; eight external stakeholders | World Café notes and posters; model-development records; undergraduate scenario tasks; master's MCDA reports; stakeholder evaluation materials | Descriptive process analysis of the four implementation stages and synthesis of participant activities and outputs | Four-stage living-lab implementation process and description of how each participant group contributed |
| RQ2. | 196 undergraduate students in 98 pairs; 30 master's students in 15 pairs | 294 undergraduate scenario responses; model simulation outputs; master's scenario reports and MCDA evaluations | Descriptive quantitative analysis of scenario outputs; qualitative review of explicit references to system relationships, emissions, costs, performance indicators and trade-offs | Evidence of system interpretation and trade-off recognition in undergraduate and master's outputs |
| RQ3. | 15 master's student groups; eight external stakeholders | Master's preferred scenarios and accompanying written justifications; stakeholder scenario evaluations | Descriptive comparison of preferred scenarios and the decision priorities reflected in the accompanying justifications and stakeholder evaluations | Comparison of environmental, economic, technical and implementation-oriented decision priorities; 12 of 15 master's groups were classified as environmentally oriented based on their preferred scenario and accompanying justification |
Methodological overview linking research questions, participant groups, data sources, analysis and outputs.
The framework was used to structure the analysis rather than to calculate a composite measure of educational or system performance. Table 1 presents this analytical framework by linking each research question to the corresponding participant groups, data sources and instruments, analysis methods, and main outputs.
2.6 Ethics statement
The research protocol was reviewed and approved by the Research Ethics Committee of Riga Technical University (Application No. 04000-10.1-e/2; Decision No. 04000-10.2.3-e/13). The application was considered at the Committee meeting on 26 January 2026, and the formal decision letter was issued on 18 February 2026.
Participation was voluntary, and informed consent was obtained electronically from the student participants through the RTU system. The professional and external stakeholder participants received information about the study and provided informed consent before participation. No minors were involved. Participants were informed that they could decline or withdraw from participation at any time without negative consequences. Data were anonymised or pseudonymised and processed in accordance with the General Data Protection Regulation, Latvian data-protection requirements, and RTU internal data-protection procedures. The study involved no physical risks and only minimal psychological risks related to expressing opinions and reflecting on energy consumption and CO₂-footprint reduction. The research was conducted in accordance with the principles of the Declaration of Helsinki.
3 Results
The results are organised according to the three research questions. Section 3.1 describes how the district-heating digital-twin-based modelling environment was integrated into the four-stage living-lab process. Section 3.2 examines evidence of system interpretation and trade-off recognition in undergraduate and master's student outputs. Section 3.3 compares the decision priorities of master's student groups and external stakeholders. Section 3.4 synthesises the contributions of the four participant groups to the overall process. The study was conducted from February 2025 to March 2026 at Riga Technical University, Latvia.
3.1 Integration of the digital-twin-based modelling environment into the multilevel living lab
The living lab was implemented through a structured sequence involving co-creation, large-scale educational application, analytical evaluation, and stakeholder assessment. Each stage contributed to refining the modelling environment and examining its use from different participant perspectives.
The first phase of the co-creation work involved professionals from the selected municipalities and energy companies. They discussed and defined the key components and limitations of the system and identified potential development paths. The World Café technique was used to collect domain knowledge regarding system dynamics, limitations and planful considerations. Eight tables worked intensively in conversations, and the outcomes were summarised and concretised into four system concepts for the development of the modelling framework, including the system boundaries, possible technological solutions, and performance indicators such as emissions, costs, and system performance.
Participants emphasised the importance of incorporating demand variability, renewable energy integration, electrification pathways and thermal storage. These elements were subsequently incorporated into the digital model and served as the basis for scenario development in the later stages. Practitioner input informed the model boundaries, technological options, key performance indicators, and scenario parameters, including demand variability, renewable-energy integration, electrification pathways, and thermal storage.
The second stage examined the implementation of the modelling activity with a relatively large undergraduate cohort. A total of 196 undergraduate students worked in 98 pairs and produced 294 scenario responses. The students varied selected technical and operational parameters and considered the resulting changes in emissions, costs, and system performance. Their written outputs were examined for explicit references to model outputs, relationships between system components, and environmental and economic trade-offs.
The modelling activity was implemented with 196 undergraduate students working in 98 pairs, showing that the activity could be delivered to a relatively large student cohort within the examined institutional setting. However, usability was not assessed using a validated instrument, and the findings should therefore be interpreted as evidence of implementation rather than demonstrated usability or effectiveness.
3.2 Evidence of system interpretation and trade-off recognition
The undergraduate responses showed different frequencies of environmental, economic, technical, and integrative elements. Economic effects were identified in 250 of the 294 responses (85.0%), while technical or system relationships were present in 225 responses (76.5%). Trade-offs between competing outcomes were explicitly recognised in 160 responses (54.4%), and environmental effects were referred to in 150 responses (51.0%). Integrated recommendations combining at least two environmental, economic, or technical dimensions were identified in 110 responses (37.4%). Because the categories were non-exclusive, individual responses could contribute to more than one category. These findings provide descriptive evidence of how students engaged with system relationships and sustainability trade-offs within the assigned activity; they do not represent measured learning gains or formally assessed competence.
Representative paraphrased examples from the student responses illustrated these patterns. For example, one response linked increased renewable-energy use with lower emissions but higher investment costs, while another identified thermal storage as improving system flexibility while requiring additional investment. Such responses illustrate the types of system relationships and trade-offs captured by the descriptive categories. The frequency of the five descriptive categories identified in the 294 undergraduate scenario responses is summarised in Table 2.
Table 2
| Category identified in undergraduate responses | n | % |
|---|---|---|
| Economic effects | 250 | 85.0 |
| Technical/system relationships | 225 | 76.5 |
| Trade-off recognition | 160 | 54.4 |
| Environmental effects | 150 | 51.0 |
| Integrated recommendations | 110 | 37.4 |
Descriptive categories identified in the 294 undergraduate scenario responses.
Categories were non-exclusive; therefore, percentages do not sum to 100%.
At the master's level, 30 students worked in 15 pairs. Each pair developed an environmentally oriented scenario, a cost-oriented scenario, and a balanced scenario. These three alternatives were ranked using TOPSIS, after which each pair selected and justified a final preferred scenario. Twelve of the 15 groups (80%) were classified as environmentally oriented because their written justification identified emissions reduction or broader environmental performance as a principal reason for selecting the preferred scenario; the selected scenario also showed the strongest environmental performance among the alternatives considered.
This indicates a common decision orientation within this cohort but should not be interpreted as evidence of learning effectiveness.
3.3 Comparison of master's student and stakeholder priorities
The master's student groups and external stakeholders differed in the decision priorities reflected in their evaluations. The master's student groups used TOPSIS for the multi-criteria comparison of scenarios, with equal weights assigned to all indicators. The external stakeholders did not repeat the TOPSIS procedure; instead, they evaluated the proposed scenarios and model outputs according to environmental performance, investment cost, technical feasibility, system reliability, and practical implementation considerations. Twelve of the 15 master's student groups (80%) prioritised emissions reduction or broader environmental performance when selecting their preferred scenario.
The eight external stakeholders comprised two ministry representatives, two municipal representatives, and four energy-company representatives. The discussion notes suggested different priority patterns across organisational roles. Municipal representatives referred mainly to levelised cost of energy and implementation feasibility, while energy-company representatives focused on energy expenses, levelised cost of energy, reliability, and practical implementation. Ministry representatives placed comparatively greater emphasis on emissions reduction and environmental performance. These findings are interpreted as qualitative descriptive patterns rather than quantitatively established dominant priorities. Table 3 summarises the stakeholder composition, the priority categories identified from the discussion notes, and the supporting evidence for each stakeholder role.
Table 3
| Stakeholder role | n | Identified priority categories | Supporting evidence |
|---|---|---|---|
| Ministry representatives | 2 | Emissions reduction; environmental performance | Discussion notes referring to environmental and emissions-related considerations |
| Municipal representatives | 2 | Levelised cost of energy; implementation feasibility | Discussion notes referring to cost and feasibility considerations |
| Energy-company representatives | 4 | Energy expenses; levelised cost of energy; reliability; practical implementation | Discussion notes referring to operating costs, reliability and implementation issues |
Stakeholder roles, identified priority categories, and supporting evidence.
The identified priority categories were derived descriptively from the stakeholder discussion notes. Reliable frequency analysis was not possible because of the small sample and the non-standardised nature of the notes. The findings should therefore be interpreted as qualitative descriptive interpretations rather than quantitatively established dominant priorities.
Figure 3 summarises these observed decision priorities across the master's student groups and external stakeholder roles. The comparison is descriptive and should not be interpreted as an inferential comparison between equivalent participant groups.
Figure 3
For the master's student groups, TOPSIS was applied using equal weights for all indicators. External stakeholder priorities were derived from their separate evaluation of the proposed scenarios and model outputs. Twelve of the 15 master's student groups (80%) prioritised emissions reduction or broader environmental performance. Among the external stakeholders, municipal representatives emphasised levelised cost of energy, energy-company managers emphasised energy expenses and levelised cost of energy, and ministry representatives emphasised emissions reduction. The figure presents descriptive findings rather than inferential group differences.
The comparison should nevertheless be interpreted as exploratory because the master's student groups and external stakeholders followed different evaluation procedures. The master's student groups applied TOPSIS to compare their scenarios using the same set of indicators and equal indicator weights. The external stakeholders did not repeat the TOPSIS calculation; instead, they assessed the proposed scenarios and model outputs according to environmental performance, investment cost, technical feasibility, system reliability, and practical implementation considerations. Therefore, the stakeholder emphasis on particular criteria reflects the priorities expressed in their evaluations rather than different numerical weights within the TOPSIS calculation. The findings consequently represent differences in observed decision priorities rather than a controlled comparison between equivalent participant groups and evaluation procedures.
3.4 Synthesis of the four-stage process
The four-stage process connected different participant groups with different functions around the same district-heating modelling environment. In Stage 1, 24 professionals contributed practitioner knowledge concerning model boundaries, system variables, technological pathways, and relevant performance indicators. In Stage 2, 196 undergraduate students used the model in a large-cohort scenario activity and produced written responses concerning system relationships and trade-offs. In Stage 3, 30 master's students working in 15 pairs applied multi-criteria analysis to compare and justify alternative scenarios. In Stage 4, eight external stakeholders assessed the practical relevance of the scenarios and emphasised investment cost, feasibility, reliability, and implementation considerations.
The process therefore linked practitioner-informed model development, exploratory educational use, structured analytical evaluation, and external stakeholder assessment. These activities demonstrate how one modelling environment was incorporated across several stages and participant groups within the present institutional case.
4 Discussion
This case study examined the integration of a district-heating digital model into a four-stage living-lab process involving practitioners, undergraduate students, master's students, and external stakeholders. The findings are considered in relation to the three issues identified in the Introduction: integration across educational levels and a relatively large cohort, evidence of system interpretation and trade-off recognition in participant outputs, and differences between learner and stakeholder decision priorities.
Regarding RQ1, the study documents how one digital modelling environment was used sequentially for practitioner co-creation, undergraduate scenario exploration, master's-level multi-criteria analysis, and external stakeholder assessment. The four stages were not identical applications of the model but represented distinct forms of participation. Practitioners informed the model boundaries and variables, undergraduate students explored scenario relationships, master's groups conducted structured comparisons, and external stakeholders assessed practical relevance and implementation considerations. This case therefore extends living-lab implementation beyond a single participant group or educational level. The inclusion of 196 undergraduate students also demonstrates feasibility within a relatively large course cohort. However, it does not establish scalability across institutions, programmes, delivery formats, or repeated cohorts.
This qualification is consistent with the literature on higher-education living labs. , , and emphasise that living-lab outcomes depend not only on the real-world setting but also on institutional integration, pedagogical design, stakeholder relationships, facilitation, and continuity. The present case illustrates one way of organising these elements around a common modelling environment, but further comparative implementation is needed before broader institutional conclusions can be drawn.
Regarding RQ2, undergraduate responses contained references to relationships among system parameters, emissions, costs, and system performance, while master's reports included more structured multi-criteria comparisons and justifications. These outputs suggest that the model provided a setting in which participants could engage with system relationships and sustainability trade-offs. Similar potential has been reported in digital-twin education studies, where simulation environments have been used to connect technical representations with experiential and analytical tasks (; ; ). Nevertheless, the present findings concern evidence visible in participant-produced work rather than measured learning gains. No control group, pre-test/post-test design, validated learning instrument, or independent competence assessment was used.
The difference between undergraduate and master's outputs should also be interpreted cautiously. The master's students completed a more advanced assignment requiring multi-criteria analysis and explicit justification, whereas the undergraduate task was exploratory. The more structured reasoning in the master's reports may therefore reflect differences in task design, educational level, prior knowledge, or assessment expectations rather than an effect attributable solely to use of the modelling environment.
Regarding RQ3, 12 of the 15 master's groups selected scenarios associated primarily with environmental performance, whereas the external stakeholder evaluations placed greater emphasis on investment cost, feasibility, reliability, and implementation considerations. This result is consistent with previous evidence that students, professionals, employers, and other stakeholder groups may evaluate sustainability-related priorities differently. found that student evaluations did not strongly predict professional preferences in sustainable-design practices. similarly identified differences among students, educators, and employers in their perceptions of sustainability competences, while showed that practitioner assessments of sustainable projects often give greater weight to cost, experience, project maturity, and technology transfer. One possible interpretation is that the master's students approached the task primarily as an analytical sustainability exercise, whereas the external stakeholders evaluated the scenarios from organisational roles that also involve responsibility for investment, implementation feasibility, system reliability, and operational constraints.
In the present study, the digital model provided a shared object through which these differing priorities could be made visible. Its contribution was therefore not to eliminate disagreement or determine one objectively preferred scenario, but to support structured comparison of alternatives and discussion of the criteria applied by different participant groups. Because the student and stakeholder evaluations were conducted in different contexts and were not fully equivalent, this finding should be regarded as an exploratory comparison rather than a controlled test of group differences.
Figure 1 should therefore be interpreted as a conceptual synthesis derived from the present case, not as a validated framework with established transferability. It identifies possible connections among educational, analytical, and stakeholder-oriented uses of digital modelling. The framework may provide a basis for testing in other institutional and sectoral contexts, but the present study does not demonstrate broader scalability or effectiveness.
Taken together, the findings suggest that digital modelling can serve different functions within a living-lab process: as a common representation for exploratory learning, as a basis for structured multi-criteria analysis, and as a means of making different stakeholder priorities visible. These functions are represented conceptually in Figure 1. Their transferability to other institutional or sectoral settings remains to be tested.
This study has several limitations. First, it was conducted at a single university using one district-heating case, which limits transferability to other institutions, educational settings, and sustainability domains. Second, the study did not include a control group, a pre-test/post-test design, or a validated learning assessment instrument. It is therefore not possible to determine whether the modelling environment improved learning outcomes or whether it was more effective than conventional teaching methods. The student outputs provide descriptive evidence of systems reasoning and trade-off recognition within the assigned tasks rather than a direct measurement of educational effectiveness.
Third, the undergraduate analysis relied on written work produced during a course activity. The responses may have been influenced by task instructions, group interaction, task expectations, and instructor support. Fourth, the external stakeholder sample consisted of eight participants, and the findings may not represent the wider range of actors involved in district-heating transitions. In addition, the master's student groups and external stakeholders followed different evaluation procedures: the master's groups used TOPSIS with equal indicator weights, whereas the external stakeholders separately evaluated the scenarios and model outputs according to environmental performance, investment cost, technical feasibility, system reliability, and practical implementation considerations. This limits direct comparison between the groups.
The digital model was used as a scenario-based representation of the district-heating system rather than as a fully operational digital twin with continuous real-time data exchange. Model development was informed by practitioner input and stakeholder review, but an independent technical validation against a separate operational dataset was not conducted. The findings should therefore be interpreted in relation to the scenario-based modelling environment used in this case. In addition, the analyses were descriptive and no inferential statistical tests were performed. Future studies should validate model outputs against operational data, incorporate real-time data streams where feasible, and use larger samples and statistical comparison methods when the research design permits.
Fifth, the digital modelling environment was based on scenario simulations rather than fully integrated real-time operational data. Sixth, the study did not include inferential statistical testing, and the analysis was primarily descriptive. Finally, the qualitative analysis relied on descriptive review and researcher interpretation of participant-produced material rather than on a formal qualitative coding procedure. Future studies could strengthen this component by applying a predefined coding framework, involving multiple coders where possible, and reporting inter-coder agreement. Future studies should repeat the intervention across institutions and cohorts, use validated learning and usability measures, apply equivalent evaluation procedures across participant groups, and examine whether incorporating real-time operational data changes the educational and decision-support value of the modelling environment.
Future research should explore applications in different contexts, integrate real-time data, and examine their long-term impacts. Addressing these limitations would further strengthen the evidence base for the scalability and long-term effectiveness of our proposed approach.
Linking living labs and digital twins may address global sustainability challenges by combining an educational dimension with model development and evaluation using real-world phenomena, while advancing the use of analytical models.
5 Conclusions
This study shows how a common digital modelling environment can connect different forms of participation within a multilevel living-lab process, from practitioner-informed model development and exploratory student use to structured decision analysis and external stakeholder assessment. The value of this approach lies not only in the use of the digital model itself, but in its ability to provide a shared representation of a complex sustainability problem that can be approached at different levels of knowledge, responsibility, and decision-making.
The findings suggest that such an environment can support engagement with relationships among technical, environmental, and economic aspects of an energy system and make sustainability trade-offs more explicit. At the same time, the differences observed between master's student and external stakeholder evaluations show that a common modelling environment does not necessarily lead to common priorities. Rather, it can make contrasting perspectives visible and thereby provide a basis for discussion between educational and professional actors.
The study therefore contributes to the literature on living labs and digital twins by illustrating how a common digital modelling environment can be embedded within a multilevel living lab and used across several educational levels and stakeholder groups. While previous research has often examined living labs or digital-twin-supported learning within individual courses, participant groups, or specialised settings, this case demonstrates their integration across multiple stages of participation. It also adds to the limited evidence on how learner and practitioner groups may interpret the same digitally represented sustainability challenge from different decision perspectives.
These conclusions should be interpreted within the scope of an exploratory single-institution case. The study does not establish educational effectiveness, general scalability, or superiority over conventional teaching or decision-support approaches. Further research should therefore test the approach across institutions and cohorts, use validated learning and usability measures, apply comparable evaluation procedures where direct group comparison is intended, and investigate whether integration of real-time operational data strengthens its educational and decision-support value.
Statements
Ethics statement
The research protocol was reviewed and approved by the Research Ethics Committee of Riga Technical University (Application No. 04000-10.1-e/2; Decision No. 04000-10.2.3-e/13). The application was considered at the Committee meeting on 26 January 2026, and the formal decision letter was issued on 18 February 2026.
Participation was voluntary, and informed consent was obtained electronically from the student participants through the RTU system. The professional and external stakeholder participants received information about the study and provided informed consent before participation. No minors were involved. Participants were informed that they could decline or withdraw from participation at any time without negative consequences. Data were anonymised or pseudonymised and processed in accordance with the General Data Protection Regulation, Latvian data-protection requirements, and RTU internal data-protection procedures. The study involved no physical risks and only minimal psychological risks related to expressing opinions and reflecting on energy consumption and CO₂-footprint reduction. The research was conducted in accordance with the principles of the Declaration of Helsinki.
Author contributions
DP: Project administration, Writing – review & editing, Writing – original draft, Investigation, Funding acquisition. JP: Data curation, Methodology, Writing – review & editing, Writing – original draft. UG: Writing – review & editing, Writing – original draft, Visualization. AB: Validation, Writing – review & editing, Writing – original draft. DB: Writing – original draft, Supervision, Writing – review & editing.
Funding
The author(s) declared that financial support was received for this work and/or its publication. This project has received funding from the European Union’s Horizon Europe Framework Programme under Grant Agreement No. 101139527.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
References
1
AghoM. O.Eyo-UdoN. L.OnukwuluE. C.SuleA. K.AzubuikeC. (2025). Digital twin technology for real-time monitoring of energy supply chains. Int. J. Res. Innov. Appl. Sci.9, 564–592. 10.51584/IJRIAS.2024.912049
2
AlyamaniR.LongS.NurunnabiM. (2021). Evaluating decision making in sustainable project selection between literature and practice. Sustainability13, 8216. 10.3390/su13158216
3
AttahR. U.GarbaB. M. P.Gil-OzoudehI.IwuanyanwuO. (2024). Evaluating strategic technology partnerships: providing conceptual insights into their role in corporate strategy and technological innovation. Int. J. Front. Sci. Technol. Res.7, 77–89. 10.53294/ijfstr.2024.7.2.0058
4
BaratS.KulkarniV.ClarkT.BarnB. (2022). “Digital twin as risk-free experimentation aid for techno-socio-economic systems,” in MODELS ‘22: Proceedings of the 25th International Conference on Model Driven Engineering Languages and Systems (New York, NY: Association for Computing Machinery), 66–75. 10.1145/3550355.3552409
5
BasseyK. E.Opoku-BoatengJ.AntwiB. O.NtiakohA.JulietA. R. (2024). Digital twin technology for renewable energy microgrids. Eng. Sci. Technol. J.5, 2248–2272. 10.51594/estj.v5i7.1319
6
BaumberA. (2022). Transforming sustainability education through transdisciplinary practice. Environ. Dev. Sustain.24, 7622–7639. 10.1007/s10668-021-01731-3
7
BerberiA.BeaudoinC.McPheeC.GuayJ.BronsonK.NguyenV. M. (2026). Enablers, barriers, and future considerations for living lab effectiveness in environmental and agricultural sustainability transitions: a review of studies evaluating living labs. Local Environ.31, 1–19. 10.1080/13549839.2023.2238750
8
BouwmaI.WigboldusS.PottersJ.SelnesT.Van RooijS.WesterinkJ. (2022). Sustainability transitions and the contribution of living labs: a framework to assess collective capabilities and contextual performance. Sustainability14, 15628. 10.3390/su142315628
9
Cazorla-MonteroA.De Los Ríos-CarmenadoI.PastenJ. (2019). Sustainable development planning: master’s based on a project-based learning approach. Sustainability11, 6384. 10.3390/su11226384
10
De GooyertV.AwanA.GürsanC.SwennenhuisF.JanipourZ.GonellaS. (2024). Building on and contributing to sustainability transitions research with qualitative system dynamics. Sustain. Sci.19, 1949–1962. 10.1007/s11625-024-01548-9
11
DemssieY. N.BiemansH. J. A.WesselinkR.MulderM. (2023). Fostering students’ systems thinking competence for sustainability by using multiple real-world learning approaches. Environ. Educ. Res.29, 261–286. 10.1080/13504622.2022.2141692
12
FaludiJ.YiuF.SrourO.KamareddineR.AliO.MecannaS. (2019). Do student trials predict what professionals value in sustainable design practices?J. Mech. Des.141, 102001. 10.1115/1.4043200
13
GrievesM. (2024). “Digital twins and their role in reengineering engineering education,” in Digital Twin: Fundamentals and Applications, eds. SabriS.AlexandridisK.LeeN. (Cham: Springer Nature Switzerland), 237–261. 10.1007/978-3-031-67778-6_11
14
GunarathnaC.YangR.Wijeratne MudiyanselageP.AmarasingheG.SamarasinghalageT.WeerasingheR. P. N.et al (2024). Project-based learning for proactive skills development of postgraduate students in solar energy building design digitalisation. Smart Sustain. Built Environ.13, 828–855. 10.1108/SASBE-08-2022-0173
15
HadfieldP.SharpD.PigeonJ.RavenR. (2025). Governing university living labs for sustainability transformations: insights from 18 international case studies. Sustain. Sci.20, 1753–1774. 10.1007/s11625-025-01690-y
16
HerthA.VerburgR.BlokK. (2025a). How can campus living labs thrive to reach sustainable solutions? Clean. Prod. Lett.8, 100078. 10.1016/j.clpl.2024.100078
17
HerthA.VerburgR.BlokK. (2025b). The innovation power of living labs to enable sustainability transitions: challenges and opportunities of on-campus initiatives. Creat. Innov. Manag.34, 427–442. 10.1111/caim.12649
18
HuijbenJ. C. C. M.Van Den BeemtA. A. J.WieczorekA. J.Van MarionM. H. (2022). Networked learning to educate future energy transition professionals: results from a case study. Eur. J. Eng. Educ.47, 446–466. 10.1080/03043797.2021.1978403
19
HwangC.-L.LaiY.-J.LiuT.-Y. (1993). A new approach for multiple objective decision making. Comput. Oper. Res.20, 889–899. 10.1016/0305-0548(93)90109-V
20
JinC.LiuR. (2024). An evaluation of capabilities, benefits, and challenges of developing digital twin models for sustainable development. Hum. Factors Des. Eng. Comput.159, 1641–1651. 10.54941/ahfe1005732
21
KaranamS. A. K.HartmanN. W. (2025). A systematic review of digital twin and virtual learning environments for smart manufacturing education. Manuf. Lett.44, 1597–1608. 10.1016/j.mfglet.2025.06.179
22
KonstantinidisE. I.PetsaniD.BamidisP. D. (2021). Teaching university students co-creation and living lab methodologies through experiential learning activities and preparing them for RRI. Health Informatics J.27, 1460458221991204. 10.1177/1460458221991204
23
LiljaniemiA.PaavilainenH. (2020). Using digital twin technology in engineering education—course concept to explore benefits and barriers. Open Eng10, 377–385. 10.1515/eng-2020-0040
24
LöhrK.WeinhardtM.SieberS. (2020). The “world café” as a participatory method for collecting qualitative data. Int. J. Qual. Methods19, 1609406920916976. 10.1177/1609406920916976
25
MolinariM.Anund VogelJ.RolandoD.LundqvistP. (2023). Using living labs to tackle innovation bottlenecks: the KTH live-in lab case study. Appl. Energy338, 120877. 10.1016/j.apenergy.2023.120877
26
MonekG. D.FischerS. (2024). Expert twin: a digital twin with an integrated fuzzy-based decision-making module. Decis. Mak. Appl. Manag. Eng.8, 1–21. 10.31181/dmame8120251181
27
Moreno-CelyA.Cuajera-NahuiD.Escobar-VasquezC. G.VanwingT.Tapia-PonceN. (2021). Breaking monologues in collaborative research: bridging knowledge systems through a listening-based dialogue of wisdom approach. Sustain. Sci.16, 919–931. 10.1007/s11625-021-00937-8
28
OmranyH.Al-ObaidiK. M.GhaffarianhoseiniA.ChangR.-D.ParkC.RahimianF. (2026). Digital twin technology for education, training and learning in construction industry: implications for research and practice. Eng. Constr. Archit. Manag.33, 1836–1870. 10.1108/ECAM-10-2024-1376
29
Pando-GarciaJ.García-MerinoJ. D.Periañez-CañadillasI.Luengo-ValderreyM. J. (2026). Analysing multi-stakeholder perceptions of sustainability education competences. Is there a shared view?Int. Rev. Public Nonprofit Mark.10.1007/s12208-026-00486-9
30
PathakD.UpadhyayD. (2024). Integration of digital twins technologies in education for experiential learning: benefits and challenges. Int. Res. J. Adv. Eng. Hub2, 442–449. 10.47392/irjaeh.2024.0064
31
Pinon FischerO. J.SabriS.ChenY. (2024). “Fundamentals of digital twins, modeling approaches, and governance,” in Digital Twin: Fundamentals and Applications, eds. SabriS.AlexandridisK.LeeN. (Cham: Springer Nature Switzerland), 13–29. 10.1007/978-3-031-67778-6_2
32
Ratnam-LimC. T. L.TanK. H. K. (2015). Large-scale implementation of formative assessment practices in an examination-oriented culture. Assess. Educ. Princ. Policy Pract.22, 61–78. 10.1080/0969594X.2014.1001319
33
Rodríguez-ChuecaJ.Molina-GarcíaA.García-ArandaC.PérezJ.RodríguezE. (2019). Understanding sustainability and the circular economy through flipped classroom and challenge-based learning: an innovative experience in engineering education in Spain. Environ. Educ. Res.26, 238–252. 10.1080/13504622.2019.1705965
34
RovatiL.GaryP. J.CubroE.DongY.KilickayaO.SchulteP. J.et al (2024). Development and usability testing of a patient digital twin for critical care education: a mixed methods study. Front. Med.10, 1336897. 10.3389/fmed.2023.1336897
35
SabriS.AlexandridisK.LeeN. (2024). “Introduction to digital twins,” in Digital Twin: Fundamentals and Applications, eds. SabriS.AlexandridisK.LeeN. (Cham: Springer Nature Switzerland), 1–12. 10.1007/978-3-031-67778-6_1
36
SpyrouO.Ariza-SentísM.VélezS. (2025). Enhancing education in agriculture via XR-based digital twins: a novel approach for the next generation. Appl. Syst. Innov.8, 38. 10.3390/asi8020038
37
StuckrathC.Rosales-CarreónJ.WorrellE. (2025). Conceptualisation of campus living labs for the sustainability transition: an integrative literature review. Environ. Dev.54, 101143. 10.1016/j.envdev.2025.101143
38
TercanliH.JongbloedB. (2022). A systematic review of the literature on living labs in higher education institutions: potentials and constraints. Sustainability14, 12234. 10.3390/su141912234
39
TranT.NguyenQ.LuuT.TranM.KuaJ.HoangT.et al (2025). Empowering robotic training with kinesthetic learning and digital twins in human-centric industrial systems. J. Ind. Inf. Integr.43, 100743. 10.1016/j.jii.2024.100743
40
TrischlerJ.SvenssonP. O.WilliamsH.WikströmF. (2023). Citizens as an innovation source in sustainability transitions – linking the directionality of innovations with the locus of the problem in transformative innovation policy. Public Manag. Rev.25, 2093–2115. 10.1080/14719037.2022.2062041
41
UhlenkampJ.-F.Baalsrud HaugeJ.BrodaE.LütjenM.FreitagM.ThobenK.-D. (2022). Digital twins: a maturity model for their classification and evaluation. IEEE. Access.10, 69605–69635. 10.1109/ACCESS.2022.3186353
42
Van der WeeM. L. E.TassoneV. C.WalsA. E. J.TroxlerP. (2024). Characteristics and challenges of teaching and learning in sustainability-oriented living labs within higher education: a literature review. Int. J. Sustain. High. Educ.25, 255–277. 10.1108/IJSHE-10-2023-0465
43
Van GeenhuizenM. (2018). A framework for the evaluation of living labs as boundary spanners in innovation. Environ. Plan. C Polit. Space36, 1280–1298. 10.1177/2399654417753623
44
VerhoefL. A.BossertM.NewmanJ.FerrazF.RobinsonZ. P.AgarwalaY.et al (2020). “Towards a learning system for university campuses as living labs for sustainability,” in Universities as Living Labs for Sustainable Development: Supporting the Implementation of the Sustainable Development Goals, eds. Leal FilhoW.Lange SalviaA.PretoriusR. W.Londero BrandliL.ManolasE.AlvesF.AzeiteiroU.RogersJ.ShielC.Do PacoA. (Cham: Springer), 135–149. 10.1007/978-3-030-15604-6_9
45
WeichbrothP.JandyK.ZuradaJ. (2024). Toward sustainable development: exploring the value and benefits of digital twins. Telecom5, 774–791. 10.3390/telecom5030039
Summary
Keywords
digital twin, district heating systems, energy system modeling, experiential learning, living lab, multi-criteria decision-making, sustainability transitions
Citation
Paule D, Pubule J, Gabranova U, Blumberga A and Blumberga D (2026) A multilevel living lab using a digital-twin-based modelling environment for sustainability education and decision support. Front. Educ. 11:1863292. doi: 10.3389/feduc.2026.1863292
Received
23 April 2026
Revised
24 August 2026
Accepted
04 September 2026
Published
02 October 2026
Volume
11 - 2026
Edited by
Abena Agyeiwaa Obiri-Yeboah, Kumasi Polytechnic, Ghana
Reviewed by
Peter Troxler, Rotterdam University of Applied Sciences, Netherlands
Kadek Dwi Hendratma Gunawan, Sebelas Maret University, Indonesia
Updates
Copyright
© 2026 Paule, Pubule, Gabranova, Blumberga and Blumberga.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Dace Paule dace.paule@rtu.lv
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.