Review Reports - A Dynamic Digital Twin Framework for Sustainable Facility Management in a Smart Campus: A Case Study of Chiang Mai University

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript titled "A Dynamic Digital Twin Framework for Sustainable Facility Management in a Smart Campus: A Case Study of Chiang Mai University" presents a comprehensive and well-structured approach to enhancing sustainable facility management in a smart campus environment. The study’s integration of 3D spatial modeling, real-time environmental and energy sensor data, and multiscale dashboard visualizations is innovative and relevant to the growing need for sustainability in higher education institutions. The use of stakeholder-driven requirements, combined with a four-phase methodology, provides helpful information regarding the practical deployment of digital twin technology for facility management.

However, there are several points that could be clarified or expanded upon to further strengthen the manuscript.

1. Clarity of Methodology:
The four-phase development methodology is well-articulated, but it would be helpful to provide additional details on the specific stakeholder consultation process. For instance, how were key stakeholders (e.g., campus administrators, technicians, and students) involved in each phase? Were there any challenges in aligning diverse stakeholder needs, and how were these addressed? Further elaboration on the feedback loops and how these shaped the platform's final design would provide a more comprehensive understanding of the design and development process.

The manuscript mentions the deployment of IoT sensors using NB-IoT and LoRaWAN protocols, but it would be beneficial to provide additional information on the selection of these protocols. Specifically, why were these particular communication protocols chosen over others? What were the main challenges encountered during the deployment of the sensors, and how were they mitigated? Furthermore, an explanation of the types of sensors used (e.g., temperature, humidity, and energy meters) and the accuracy or reliability of these measurements would add to the manuscript’s depth.

The user testing described in the manuscript demonstrates the platform's effectiveness in supporting interactive monitoring. However, a more detailed analysis of user feedback would be valuable. For example, which aspects of the platform were most appreciated by campus users, and were there any features that users found difficult to navigate or less useful? Additionally, what specific improvements were suggested during the testing phase, and have any of these been incorporated into the final design? This type of data would provide readers with a clearer understanding of the platform’s usability and its potential for broader adoption.

While the manuscript emphasizes energy management as the top priority, more information on the platform’s impact on energy consumption would strengthen the argument for its effectiveness in promoting sustainability. Were there any specific energy savings or efficiencies observed during the pilot phase? Are there any quantitative metrics or data available to demonstrate how the system has helped optimize energy use in the Faculty of Engineering? Providing such data would offer concrete evidence of the platform’s success in achieving sustainability goals.

The manuscript highlights the modular and hardware-agnostic architecture of the system, which allows for future extensions such as occupancy tracking, water monitoring, and automated control systems. While this is an exciting prospect, more details on how these future features could be implemented would be beneficial. For example, what are the technical and logistical challenges of adding these extensions? How might the platform handle the increased data volume and complexity as new features are integrated? A discussion of the scalability and flexibility of the system would reassure readers that the platform can evolve to meet the growing demands of smart campuses and cities.

The development of multiscale dashboards is a key strength of the platform, but more information on the design and functionality of these dashboards would be useful. How are the different levels (faculty, building, floor, room) integrated into the user interface? Are there any customization options available for users to tailor the dashboards to their specific needs? Additionally, the manuscript could benefit from including screenshots or examples of the dashboards to illustrate their usability and design. Also, make sure that the figures are of good quality and clearly readable; some of the images may need to be resized or enhanced to ensure legibility in print and digital formats.

The manuscript briefly mentions the potential for broader applications in smart cities and built environment innovation. It would be interesting to expand on this point by exploring how the digital twin framework can be adapted or scaled for use in urban environments. Are there any specific use cases or examples where this framework could be applied outside of the university setting? How might the system contribute to broader smart city goals, such as reducing carbon emissions or improving urban mobility?

The conclusion emphasizes the potential for the system to serve as a replicable and scalable model for smart cities. However, the manuscript would benefit from a more detailed discussion on the limitations of the current system and potential areas for improvement. Should future iterations of the platform address any specific technical or operational challenges? What are the next steps in terms of further research or development to enhance the system’s capabilities?

While the manuscript provides a solid foundation of references, it would benefit from the inclusion of more recent studies, particularly those published in 2024 and 2025. Incorporating these latest developments in digital twin technology, smart campuses, and sustainability would place the research in the context of the most current trends and innovations. Ensure that the references reflect the evolving landscape of facility management and smart infrastructure.

This study presents a well-thought-out and innovative approach to integrating digital twin technology for sustainable facility management. The platform’s modularity, focus on energy management, and emphasis on stakeholder involvement are key strengths. However, there are several areas where additional details, clarifications, and data would enhance the manuscript’s impact. By addressing these aspects and updating the references, the authors can further demonstrate the value of this system not only for campus management but also for broader applications in smart cities and sustainable urban development.

Author Response

Comments 1: The four-phase development methodology is well-articulated, but it would be helpful to provide additional details on the specific stakeholder consultation process. For instance, how were key stakeholders (e.g., campus administrators, technicians, and students) involved in each phase? Were there any challenges in aligning diverse stakeholder needs, and how were these addressed? Further elaboration on the feedback loops and how these shaped the platform's final design would provide a more comprehensive understanding of the design and development process.

Response 1: We agree with adding details about the processes involving stakeholders, specifying their roles in each step, the challenges encountered, and how these were addressed. Such information clearly reflects the iterative process that ultimately informed the final design. Thank you for pointing this out. Therefore, we have added information in each step of the Methodology regarding how stakeholders were involved in the experimental process. Specifically, in Step 3.1 Data Collection and Requirement Analysis,

“Added:

In designing and developing the Digital Twin Framework for Sustainable Facility Management in a Smart Campus, it is essential to address these management requirements. Regarding the data collection process, the primary stakeholders involved were the building management team, consisting of (1) building resource management personnel and (2) maintenance technicians. They provided physical information necessary for developing the digital twin system, as the framework requires building and environmental data to be integrated with numerical datasets for enhanced visualization.

A significant challenge was collecting baseline data, as certain records were not systematically maintained. Examples include tree species, their spatial locations, and utility infrastructure details. A dedicated survey team was established to collect, digitize, and organize these data into a structured digital database to address this.

” – page 5, paragraph 2-3, and line 198-208.

In Step 3.2 3D Model Creation,

“Added:

In this phase, the building management team participated by providing consultation on the appropriate level of model detail, as it directly affects the performance of the web-based visualization. Excessively detailed models would require high computational resources, negatively impacting usability. A challenge was also encountered with older buildings that lacked existing digital records; this was addressed through on-site surveys to generate the necessary models. The ICT services team also provided guidance and managed the computing environment to ensure the platform could be effectively supported.

” – page 7, paragraph 1, and line 273-279.

In Step 3.3 Sensor Deployment and Data Acquisition,

“Added:

In this phase, the building management team, particularly the maintenance technicians, participated by providing consultation on installing and integrating sensors to maximize the measurement and monitoring of energy consumption and air quality. Collaborations were established with ERDI and DustBoy to ensure data accuracy and reliability, which provided reference energy and air quality data for validation and effective comparison.

” – page 9, paragraph 7, and line 395-370.

In Step 3.4 Data Integration Algorithm Design,

“Added:

The visualization design was developed and aligned with stakeholder requirements, as the building management team emphasized the need to consolidate physical and informational data into a single platform. This integration was intended to improve usability and clearly represent the facility environment. However, challenges arose due to limited resources and available tools, leading to an initial platform design that prioritized energy management and environmental monitoring. Other functions are planned for incremental development in later stages.

” – page 9, paragraph 10, and line 381-387.

And then, we describe their engagement, and finally, we present the impact of stakeholder involvement on the final design in Section 4.1. Results of Data Collection and Requirement Analysis,
“Added:

The results of the stakeholder requirement survey directly influenced the system design. The top priority identified was visualizing physical data alongside energy and environmental information collected from sensors. Consequently, the dashboard design highlights these datasets as the primary display. The second priority was including a maintenance alert system, which necessitated data collection on electrical appliances and building equipment. The third priority, water flow tracking, presented significant challenges, as modeling and collecting data on the building's entire piping system would be resource-intensive and difficult to achieve. Therefore, this requirement was considered a potential development path for future expansion, alongside other secondary needs.

” – page 14, paragraph 3, and line 460-468.

Comments 2: The manuscript mentions the deployment of IoT sensors using NB-IoT and LoRaWAN protocols, but it would be beneficial to provide additional information on the selection of these protocols. Specifically, why were these particular communication protocols chosen over others? What were the main challenges encountered during the deployment of the sensors, and how were they mitigated? Furthermore, an explanation of the types of sensors used (e.g., temperature, humidity, and energy meters) and the accuracy or reliability of these measurements would add to the manuscript’s depth.

Response 2: Thank you for your comment and we agree that the network can support multiple protocols; however, we did not specify why NB-IoT and LoRaWAN were selected. Therefore, we aim to revise the content in Section 3.3 to address this “The sensor network architecture was structured around eight functional domains identified during the system requirements phase. A hybrid communication strategy combining NB-IoT and LoRaWAN was adopted, as conventional Wi-Fi proved unreli-able due to high interference and spectrum congestion on campus. These protocols were selected for their long-range transmission, intense signal penetration, and low power consumption. Alternatives such as Zigbee, Z-Wave, and Bluetooth Low Energy were considered but dismissed due to limited coverage. At the same time, Sigfox was excluded because of bandwidth constraints and dependence on external operators. Deployment challenges, including signal attenuation and sensor interoperability, were mitigated through optimized gateway placement, transmission calibration, and re-dundant sensor nodes.” – page 8, paragraph 2, and line 314-323.

And we also agree with adding information on the reliability of the sensor devices, specifying that “The network integrated temperature, humidity, energy, and air quality sensors, whose accuracy was validated against reference instruments, including the Dust Boy system operated by CMU’s Climate Change Data Center (CCDC). The results con-firmed reliable performance, with accuracy maintained within ±0.5 °C for temperature, ±3% RH, and ±2% for energy, ensuring high-quality data for facility management and sustainability analysis.” – page 9, paragraph 2, and line 339-344.

Comments 3: The user testing described in the manuscript demonstrates the platform's effectiveness in supporting interactive monitoring. However, a more detailed analysis of user feedback would be valuable. For example, which aspects of the platform were most appreciated by campus users, and were there any features that users found difficult to navigate or less useful? Additionally, what specific improvements were suggested during the testing phase, and have any of these been incorporated into the final design? This type of data would provide readers with a clearer understanding of the platform’s usability and its potential for broader adoption.

Response 3: Thank you and we agree with your comment, we have added further details regarding the system usability testing, highlighting what users liked, the difficulties they encountered, the improvements made, and the final design, as presented in “

User feedback indicated high satisfaction, particularly with the system's ability to provide intuitive access to campus-wide conditions and key operational data. Among the features, the most highly requested was the capability to display energy consumption at the faculty level and disaggregate it by individual buildings. The dashboard further supports detailed visualization down to the floor and room levels, enabling a clear overview of energy usage across different locations.

The second most valued feature was access to updated construction drawings, which developers and researchers had uploaded into the database. These drawings serve as a central repository, allowing users to download them when needed for maintenance or renovation purposes. System drawings were the most frequently used in practice, as they provide essential details for repairing and managing various building systems.

In addition, the dashboard incorporated a specialized feature for maintenance technicians to manage information about key electrical equipment such as lighting and air-conditioning units. These were modeled in three dimensions and linked to the latest survey data. However, integrating the 3D models with equipment attributes—including type, power consumption, maintenance schedule, and physical location—proved to be the most challenging task, requiring significant manual data collection. The system was enhanced to address this challenge, allowing maintenance technicians to input equipment information directly. Initial usability tests revealed some difficulties that were mitigated by conducting training sessions. After training, technicians demonstrated improved understanding of how to use the dashboard effectively, and the collected data were successfully transferred for use in other subsystems.” – page 22, paragraph 1-3, and line 638-659.

and “In energy management, stakeholders recommended incorporating a data export function—particularly supporting formats such as Excel—to enable further analysis and seamless integration into existing workflows. Consequently, the dashboard was upgraded to meet these needs by allowing direct download of both energy data and construction drawings through the interface.” – page 23, paragraph 2, and line 676-680.

Comments 4: While the manuscript emphasizes energy management as the top priority, more information on the platform’s impact on energy consumption would strengthen the argument for its effectiveness in promoting sustainability. Were there any specific energy savings or efficiencies observed during the pilot phase? Are there any quantitative metrics or data available to demonstrate how the system has helped optimize energy use in the Faculty of Engineering? Providing such data would offer concrete evidence of the platform’s success in achieving sustainability goals.

Response 4: Thank you for the valuable comment. We acknowledge that we did not provide concrete evidence demonstrating how the use of Digital Twin contributes to energy reduction or savings. Therefore, we would like to add Figure 18 (page 22) and information showing the trend of decreased energy consumption from 2023 to 2024, as follows:

“A comparative analysis of electricity consumption in the Faculty of Engineering between January–December 2023 and the corresponding period in 2024 demonstrates quantifiable improvements in energy efficiency during the pilot implementation phase of the platform. Although seasonal peaks in consumption remained evident in March–April and September–October, the overall electricity usage in 2024 was consistently lower than in the preceding year. Substantial reductions were observed in August (−20.9%), November (−23.3%), and December (−8.3%), resulting in an average annual decrease of approximately 9–12%. These findings suggest that the platform effectively contributed to optimizing electricity usage by minimizing unnecessary loads during periods of low occupancy and fostering greater awareness of energy consumption patterns among faculty members and students. The results provide concrete evidence that the system not only facilitates monitoring but also achieves measurable improvements in energy efficiency, thereby reinforcing its contribution to institutional sustainability objectives.” – page 22, paragraph 4, and line 663-675.

Comments 5: The manuscript highlights the modular and hardware-agnostic architecture of the system, which allows for future extensions such as occupancy tracking, water monitoring, and automated control systems. While this is an exciting prospect, more details on how these future features could be implemented would be beneficial. For example, what are the technical and logistical challenges of adding these extensions? How might the platform handle the increased data volume and complexity as new features are integrated? A discussion of the scalability and flexibility of the system would reassure readers that the platform can evolve to meet the growing demands of smart campuses and cities.

Response 5: We would like to thank you for the comment and agree to provide additional explanations regarding the integration of other system functions, how the platform can be expanded, and its flexibility. We have added the following content:

“It facilitates current energy and environmental monitoring and establishes a foundation for future system extensions.

Potential features such as occupancy tracking, water monitoring, and automated control systems can be integrated into the existing framework by leveraging the same modular data ingestion and visualization pipeline. Occupancy sensors (passive infra-red detectors, Bluetooth beacons, or Wi-Fi analytics) could be incorporated to enrich spatial intelligence and support applications such as demand-driven HVAC control and space utilization analysis. Water monitoring would require deploying flow and pressure sensors linked through NB-IoT or LoRaWAN protocols, with the resulting data mapped to the existing hierarchical dashboard. Automated control systems, such as smart lighting or adaptive HVAC, could be integrated by extending the Application Programming Interface (API) layer to support bidirectional communication, enabling the platform to monitor and actuate devices in real time.

However, technical and logistical challenges accompany these extensions. Integrating heterogeneous sensors requires careful attention to interoperability, calibration, and data validation to maintain accuracy across domains. The increased data volume and complexity—particularly when combining high-frequency occupancy streams with energy and environmental datasets—necessitate scalable cloud-based storage and optimized data-processing pipelines. Techniques like edge computing, da-ta aggregation, and anomaly detection will be critical to managing latency and reducing system load. Additionally, user training and workflow adaptation are essential to ensure facility managers and technicians can effectively interpret and act on more complex data outputs.

From a scalability perspective, the system's hardware-agnostic design and reliance on standardized APIs make it well-suited to handle incremental growth. New sensor modules or third-party data services can be integrated without major redesign, and the visualization layer can dynamically adapt to new data categories by extending the hierarchical dashboard architecture. These characteristics reassure that the platform is capable of supporting the current smart campus deployment and flexible enough to evolve alongside the growing demands of urban-scale digital twin applications.” – page 21, paragraph 2-5, and line 602-632.

Comments 6: The development of multiscale dashboards is a key strength of the platform, but more information on the design and functionality of these dashboards would be useful. How are the different levels (faculty, building, floor, room) integrated into the user interface? Are there any customization options available for users to tailor the dashboards to their specific needs? Additionally, the manuscript could benefit from including screenshots or examples of the dashboards to illustrate their usability and design. Also, make sure that the figures are of good quality and clearly readable; some of the images may need to be resized or enhanced to ensure legibility in print and digital formats.

Response 6: Thank you and your comment made us realize that we are still lacking details on essential usage and that the image quality is poor. Therefore, we agree to revise Figures 10, 12, 14, and 16 (page 16-20), and to add content in each section as follows:

In 1. Faculty-Level Dashboard, we added

“This dashboard is designed to provide a comprehensive overview of the faculty’s overall energy consumption and environmental conditions. It presents average environmental parameters—temperature, relative humidity, and PM2.5—across the entire area, with weekly, monthly, and yearly historical trends shown in graphs on the left side of the dashboard. On the right side, the dashboard displays the overall energy consumption from all 25 buildings within the faculty, including a bar chart that separates the energy consumption by individual building, along with historical trends on weekly, monthly, and yearly scales.

Additional features include a toolbar at the top left, allowing users to toggle between 3D model visualization and GIS-based mapping. It also supports the display of 2D distribution maps of temperature, relative humidity, and PM2.5 and access to an administrative panel for managing electrical equipment. This dashboard serves faculty administrators—who can use it to monitor energy usage and support energy management planning—and other users such as visitors, students, and staff, who can use it to explore the building layout and more easily locate classrooms or rooms of interest.” – page 17, paragraph 1-2, and line 531-546.

In 2. Building-Level Dashboard, we added

“The building-level dashboard adopts a similar interface to the faculty-level version. It provides average building-level environmental data on the left side, including temperature, relative humidity, and PM2.5, and overall building energy consumption on the right side. It enables users to analyze and compare energy consumption at the building scale. In addition, the dashboard incorporates bar charts that show detailed energy usage for each floor of the building. Visitors can also use this dashboard to easily navigate and identify specific floors and rooms within the building.” – page 18, paragraph 1, and line 558-564.

In 3. Floor-Level Dashboard, we added

“The floor-level dashboard also shares a consistent interface design with the faculty and building-level dashboards but introduces additional functionalities. A toolbar at the top right lets users download relevant construction floor plans, making this tool particularly useful for maintenance personnel. Moreover, the 3D models are integrated with environmental data, enabling visualization of temperature, relative humidity, and PM2.5 values through dynamic color shading. This feature helps users understand the environmental conditions of individual rooms on the floor.” – page 20, paragraph 1, and line 575-581.

In 4. Room-Level Dashboard, we added

“The room-level dashboard provides detailed information for individual rooms. On the left, it displays environmental conditions, while the right side shows the room’s total energy consumption. Energy usage is categorized into lighting, power, and air conditioning, represented in bar chart form. This break-down enables users to assess the energy impact of different types of electrical appliances within the room. Additionally, the dashboard includes a dedicated function for maintenance technicians, allowing them to manage and update appliance data in the database. This feature enhances data completeness and supports the potential development of predictive maintenance and repair alert systems in the future.” – page 21, paragraph 1, and line 593-601.

Comments 7: The manuscript briefly mentions the potential for broader applications in smart cities and built environment innovation. It would be interesting to expand on this point by exploring how the digital twin framework can be adapted or scaled for use in urban environments. Are there any specific use cases or examples where this framework could be applied outside of the university setting? How might the system contribute to broader smart city goals, such as reducing carbon emissions or improving urban mobility?

Response 7: We would like to thank you for the valuable comment. We agree to add content regarding the suggested applications for expansion to the city level, as follows:

“Beyond the university setting, the digital twin framework demonstrates strong potential for broader applications in smart cities and built environment innovation. Campuses often function as microcosms of urban environments, and the scalability of this system suggests it can be adapted for city-wide operations. For example, the multiscale dashboard architecture—capable of visualizing data at the building, floor, and room levels—can be extended to urban districts, streets, and individual public facilities.” – page 23, paragraph 4, and line 684-689.

And “Specific use cases include municipal energy management, where aggregated consumption patterns across neighborhoods could be monitored in real time to optimize demand-response strategies and support renewable energy integration. Environmental monitoring functions, already tested on the campus, could be scaled to track urban air quality, noise levels, or heat-island effects, providing valuable insights for public health and climate adaptation planning. The framework could also support smart mobility initiatives by integrating sensor data from parking facilities, electric vehicle charging stations, and traffic flows into a unified spatial dashboard.

Importantly, these applications align with broader smart city objectives such as reducing carbon emissions, improving resource efficiency, and enhancing citizen well-being. By linking building-level data to district- and city-level platforms, the system can contribute to sustainable urban governance, enabling municipalities to identify inefficiencies, implement targeted interventions, and track progress toward carbon neutrality goals. As cities worldwide pursue digital transformation strategies, the tested prototype offers a replicable, modular, and hardware-agnostic foundation for urban-scale digital twin development.” – page 23, paragraph 6-7, and line 697-712.

Comments 8: The conclusion emphasizes the potential for the system to serve as a replicable and scalable model for smart cities. However, the manuscript would benefit from a more detailed discussion on the limitations of the current system and potential areas for improvement. Should future iterations of the platform address any specific technical or operational challenges? What are the next steps in terms of further research or development to enhance the system’s capabilities?

Response 8: Thank you for your comment and we agree to provide additional explanations regarding the limitations, as well as what actions are required for further development. The details are provided as follows:

“Nonetheless, the current system has several limitations. The reliance on manual data input for specific equipment attributes, such as lighting and HVAC units, requires significant human resources and may affect long-term data accuracy. The platform's scalability will also face challenges such as occupancy tracking, water monitoring, and automated control—are integrated, potentially increasing data volume, interoperability demands, and system complexity. Operationally, user training is still essential, as initial testing showed that technical staff required structured guidance to input and interpret data effectively.

Future research and development will therefore focus on addressing these challenges. Priority areas include automating data collection through advanced sensing and edge computing, enhancing interoperability via open data standards, and strengthening data governance and cybersecurity frameworks. In addition, predictive analytics and machine learning will be incorporated to move the platform from a monitoring tool to a proactive optimization system capable of recommending interventions.

Beyond the campus context, the modular and hardware-agnostic architecture offers strong potential for replication in smart cities. By extending its functions to urban-scale applications such as energy districts, water distribution networks, and mobility systems, the platform can contribute to broader sustainability goals, including carbon reduction and climate resilience.

In conclusion, while the prototype confirms the feasibility and benefits of cam-pus-scale digital twin deployment, its ongoing evolution will require systematic attention to scalability, interoperability, and automation. Future system iterations can pro-vide a robust, replicable, and scalable model for smart city innovation and sustainable built environment management by addressing these limitations.” – page 24-25, paragraph 6-8, 1, and line 754-777.

Comments 9: While the manuscript provides a solid foundation of references, it would benefit from the inclusion of more recent studies, particularly those published in 2024 and 2025. Incorporating these latest developments in digital twin technology, smart campuses, and sustainability would place the research in the context of the most current trends and innovations. Ensure that the references reflect the evolving landscape of facility management and smart infrastructure.

Response 9: Thank you for your comment. We are pleased to add recent research to incorporate the latest advancements in Digital Twin technology, smart campuses, and sustainability. This will ensure that our work aligns with current trends and innovations. We have added seven references, as follows:

1. “The Application of Digital Twin Technology Empowered by 5G Networks and Big Data in Smart Campus” (2025)

2. “Empowering smart cities with digital twins of buildings: Applications and implementation considerations of data-driven energy modelling in building management” (2024)

3. “Digital twins for smart building at the facility management stage: a systematic review of enablers, applications and challenges” (2024)

4.“Geospatial Digital Twins for the Development of Smart Sustainable Campus Initiatives in Ecuadorian Catholic Universities” (2024)

5. “A Digital Twin Platform Based on 3D Building Models and Smart IoT for a Climate-Resilient Campus: A Case Study of National Taiwan University” (2024)

6. “Digital Twin-Empowered Cooperative Autonomous Car-sharing Services: Proof-of-Concept” (2025)

7.“Unravelling the Use of Digital Twins to Assist Decision- and Policy-Making in Smart Cities” (2024)

With the following additions to the content:

“Recent developments demonstrate the growing role of digital twins in smart campus and urban-scale applications. For instance, digital twins empowered by 5G and big data have been applied to smart campus management [1], while geospatial digital twins have been leveraged for sustainable campus initiatives in Ecuador [2]. More recently, a systematic review highlighted the key enablers and challenges of digital twins in facility management [3]. These studies underline the evolving landscape of digital twin research and position the present work within the most current trends in smart campus and smart city innovation.” – page 4, paragraph 4, and line 159-166.

“The integration of real-time sensor data and dashboard visualization in this study is consistent with emerging implementations in smart campus contexts. For example, the digital twin platform developed at Qatar University demonstrates how da-ta-driven energy modeling can empower sustainable facility management [4]. Similarly, National Taiwan University has deployed a 3D digital twin integrated with IoT to support climate-resilient campus operations [5]. These cases validate the methodological choices adopted in the present framework and demonstrate their broader applicability.” – page 7, paragraph 4, and line 293-299.

“Outside the campus environment, digital twin technologies are increasingly deployed in urban mobility and policy-making contexts. For example, recent studies have explored digital twin–empowered cooperative autonomous car-sharing services [6] and their role in optimizing vehicle routing. Others have emphasized using digital twins as decision-support tools for urban policy-making [7]. These examples highlight the scalability of the present framework and its potential contribution to cross-domain smart city objectives, including carbon reduction, climate resilience, and improved urban mobility.” – page 23, paragraph 5, and line 690-696.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The topic of the manuscript is interesting and fits the scope of the Journal. Smart campus and the associated technologies (IoT, digital twins, etc.) constitute a prolific research field. The paper requires some extra efforts to improve its quality and presentation for the prestigious journal Technologies. A set of comments are expounded hereafter.

The type of the paper should be indicated, namely, Article seems adequate.

A keyword to include would be “open-source”, if the authors agree with the suggestion.

Regarding the concept of digital twin, as the authors know, there is no a general definition but there exist different visions such as the digital model, the digital shadow and the digital twin, depending on the level of real-time data exchange between the physical asset and the virtual representation. The definition given in lines 54-55 is good but it does not show the different angles in this disruptive technology. Therefore, some statement in this sense would be desirable in the introductory section.

A common practice in scientific papers consists on placing a final paragraph in the introductory section to describe briefly the structure of the rest of the manuscript. This enhances the readability of the document.

Some figure captions have the expression “This is a figure. Schemes follow the same formatting”, which should be replaced by the figure caption descriptive of each figure.

The caption of Figure 12 appears as Figure 1.

Figures 9 to 12 show the aspect of the developed dashboards to show the results. A closer view of the part of real-time measurements and historical trends should be added in order to observe in detail the magnitudes and their values.

The mention to SDG is very adequate. However, this reviewer misses the SDG 7 which is oriented towards ensuring access to affordable, reliable, sustainable and modern energy for all.

The developed digital twin framework uses a number of open-source technologies; however, this is only mentioned in the Conclusions. The open-source feature is positive and, therefore, it is strongly suggested to emphasize it in the paper as one of its strengths. It can be supported by recent scientific publications that also deal with open-source technologies applied for sensing, data acquisition and visualization. For example, the following ones could be considered by the authors if they agree with the suggestion:

Implementation and Experimental Application of Industrial IoT Architecture Using Automation and IoT Hardware/Software. Sensors 2024, https://doi.org/10.3390/s24248074

An IoT system for a smart campus: Challenges and solutions illustrated over several real-world use cases. Internet of Things 2025, https://doi.org/10.1016/j.iot.2024.101099

Open source platform application for smart building and smart grid controls. Automation in Construction 2023, https://doi.org/10.1016/j.autcon.2022.104622

A block diagram to specifically illustrate the deployment of the sensors networks and data acquisition devices (Node MCU, Arduino, etc.) is highly desirable when describing the sensor deployment and data acquisition, in the subsection 3.3. The communication protocols must be, evidently, included in the diagram.

Figure 2 is very illustrative of the designed algorithm. Nonetheless, there is a clarification to include. Namely, there is a block which contains “Excel (CSV file)”. This reviewer supposes that it is referred to the storage of acquired data in a local file in comma separated value format. However, the reader should not have to suppose but find clear explanations in this regard.

The Conclusions should include a brief mention to the main limitations. Apart from that, this reviewer strongly agrees with the statement “The visual interface is critical in translating complex sensor data into actionable insights”, this aspect is very relevant in digital twin deployments.

The title of the sixth section should be removed since it is empty.

Author Response

Comments 1: The type of the paper should be indicated, namely, Article seems adequate.

Response 1: Thank you for your comment, and we have revised it by selecting “Article” as the type of the paper. – page 1, paragraph -, and line 1.

Comments 2: A keyword to include would be “open-source”, if the authors agree with the suggestion.

Response 2: Thank you for your comment. We agree to add the keyword “open-source” to the research, as the study makes considerable use of open-source data. – page 1, paragraph -, and line 43.

Comments 3: Regarding the concept of digital twin, as the authors know, there is no a general definition but there exist different visions such as the digital model, the digital shadow and the digital twin, depending on the level of real-time data exchange between the physical asset and the virtual representation. The definition given in lines 54-55 is good but it does not show the different angles in this disruptive technology. Therefore, some statement in this sense would be desirable in the introductory section.

Response 3: We would like to thank you for your comment and agree to add a definition of the Digital Twin in the context of sustainable urban development, as follows:

“Although there is no universally agreed-upon definition of a digital twin, the concept is often positioned along a continuum that includes the digital model (static representation), the digital shadow (one-way data flow), and the digital twin (two-way, re-al-time interaction). This study adopts the latter perspective, emphasizing its role as a dynamic system for real-time monitoring and decision support.” – page 2, paragraph 2, and line 55-59.

Comments 4: A common practice in scientific papers consists on placing a final paragraph in the introductory section to describe briefly the structure of the rest of the manuscript. This enhances the readability of the document.

Response 4: Thank you for your comment, and we agree to explain the structure of the paper to give readers an overall understanding of the content, with the following additions:

“The remainder of this research is structured as follows. Section 2 reviews related literature on smart campuses, digital twin technologies, and the integration of BIM, GIS, and IoT. Section 3 outlines the materials and methods, including requirement analysis, data collection, 3D model creation, sensor deployment, and data integration. Section 4 presents the results of system development, dashboard visualization, and prototype testing, followed by a discussion of broader implications. Finally, Section 5 concludes the study by highlighting key findings, limitations, and potential directions for future research.” – page 2-3, paragraph 7, and line 94-100.

Comments 5: Some figure captions have the expression “This is a figure. Schemes follow the same formatting”, which should be replaced by the figure caption descriptive of each figure.

Response 5: We sincerely thank you for the comment and have made corrections due to this mistake. The figure descriptions have been revised accordingly, specifically Figures 10, 12, 14, and 16 (pages 16–20).

Comments 6: The caption of Figure 12 appears as Figure 1.

Response 6: I sincerely thank you for the comment and We have revised the figure titles, and Figure 12 is now renumbered as Figure 16 (pages 20).

Comments 7: Figures 9 to 12 show the aspect of the developed dashboards to show the results. A closer view of the part of real-time measurements and historical trends should be added in order to observe in detail the magnitudes and their values.

Response 7: Thank you for your comment, and we agree to add expanded figures of each dashboard to provide clearer information, added as Figures 11, 13, 15, and 17 (pages 17–20).

Comments 8: The mention to SDG is very adequate. However, this reviewer misses the SDG 7 which is oriented towards ensuring access to affordable, reliable, sustainable and modern energy for all.

Response 8: Thank you for your comment, and we have added SDG 7 as a field directly related to the research, as indicated at – page 3, paragraph 2, and line 105.

Comments 9: The developed digital twin framework uses a number of open-source technologies; however, this is only mentioned in the Conclusions. The open-source feature is positive and, therefore, it is strongly suggested to emphasize it in the paper as one of its strengths. It can be supported by recent scientific publications that also deal with open-source technologies applied for sensing, data acquisition and visualization. For example, the following ones could be considered by the authors if they agree with the suggestion:

Response 9: Thank you for your comment, and we have incorporated the research you suggested into the following content:

[1]

Implementation and Experimental Application of Industrial IoT Architecture Using Automation and IoT Hardware/Software. Sensors 2024, https://doi.org/10.3390/s24248074

[2]

An IoT system for a smart campus: Challenges and solutions illustrated over several real-world use cases. Internet of Things 2025, https://doi.org/10.1016/j.iot.2024.101099

[3]

Open source platform application for smart building and smart grid controls. Automation in Construction 2023, https://doi.org/10.1016/j.autcon.2022.104622

“Recent studies highlight the growing role of open-source technologies in advancing digital twin and IoT applications, particularly in resource-constrained or experimental environments. For example, open platforms have been employed for industrial IoT integration [1], smart campus IoT deployments [2], and smart building and grid management [3]. Following this direction, the framework developed in this study leverages open-source tools for sensing, data acquisition, and visualization, ensuring low-cost implementation, interoperability, and replicability across different institutional and urban contexts.” – page 2, paragraph 3, and line 65-71.

And “The system is implemented using open-source tools (e.g., Firebase, QGIS, Super-Map iDesktop, and web dashboards), which reduce costs, enhance interoperability, and improve replicability across different institutional and urban contexts [1].” – page 9, paragraph 10, and line 378-380.

And “Open platforms reduce costs, enhance interoperability, and ensure replicability across institutional and urban contexts. It aligns with recent studies emphasizing the im-portance of open-source solutions in IoT-based facility management and smart infra-structure deployments [1-3].”– page 24, paragraph 4, and line 745-748.

Comments 10: A block diagram to specifically illustrate the deployment of the sensors networks and data acquisition devices (Node MCU, Arduino, etc.) is highly desirable when describing the sensor deployment and data acquisition, in the subsection 3.3. The communication protocols must be, evidently, included in the diagram.

Response 10: Thank you for your comment, and we agree with adding information on the block diagram regarding sensor data connectivity. Therefore, we have included the necessary diagram as Figure 2 (page 8) and explained the connectivity details, including the types of protocols, as follows:

“Figure 2 illustrates the detailed deployment architecture of the IoT sensor network for environmental monitoring. Multiple environmental sensors (e.g., temperature, humidity, and air-quality sensors) are connected to data acquisition devices such as NodeMCU or Arduino microcontrollers. The microcontrollers acquire and digitize the sensor readings and subsequently communicate with the NB-IoT communication module through the UART interface using AT commands. The NB-IoT module then transmits the data over the cellular network to a cloud-based MQTT server. The data is received, processed, stored, and visualized at the cloud layer in real time for facility management and monitoring purposes. The communication protocols explicitly in-volved in this architecture include UART (sensor-to-MCU), AT command protocol (MCU-to-NB-IoT), NB-IoT (module-to-cellular network), and MQTT (cloud data ex-change). This architecture provides a low-power, long-range, and reliable data acquisition solution tailored for scalable IoT deployment.” – page 8, paragraph 3, and line 326-338.

Comments 11: Figure 2 is very illustrative of the designed algorithm. Nonetheless, there is a clarification to include. Namely, there is a block which contains “Excel (CSV file)”. This reviewer supposes that it is referred to the storage of acquired data in a local file in comma separated value format. However, the reader should not have to suppose but find clear explanations in this regard.

Response 11: Thank you for your comment, and we agree, and Figure 2 is now renumbered as Figure 3. We have also added further details regarding the Excel file in – page 10, paragraph 2, and line 396.

Comments 12: The Conclusions should include a brief mention to the main limitations. Apart from that, this reviewer strongly agrees with the statement “The visual interface is critical in translating complex sensor data into actionable insights”, this aspect is very relevant in digital twin deployments.

Response 12: Thank you for your comment and we agree to provide additional explanations regarding the limitations as follows:

Comments 13: The title of the sixth section should be removed since it is empty.

Response 13: Thank you for your comment and we have removed it.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Your article makes poor reference to heat and power consumption. Table 1 contains the following:
1. Energy Management: What influences Overall energy usage?
What constitutes overall energy usage?
Compare in Literature Review with buildings in middle climates [x1].

5. Water Flow Tracking. Does this refer to hot water, cold water, or total water consumption?

The issue of pipeline anomalies is also worthy of consideration [x2], with regard to their types, causes and effects.

What are the shares of functional requirements identified by stakeholders:
In total energy consumption?
In total operating costs?

[x1] Kapalo, P.; Adamski, M. The Analysis of Heat Consumption in the Selected City. Lecture Notes in Civil Engineering, 2021. 100 LNCE pp. 158 - 165
DOI: 10.1007/978-3-030-57340-9_20

[x2] Kalenik, M.; Chalecki, M.; Wichowski, P.; Kiczko, A.; Chmielowski, K.; Świętochowska, M.; Gwoździej-Mazur J. Real values of local resistance coefficients during water flow through a pipe aerator with filling, Journal of Water and Land Development, 2023 59, pp. 174 - 182
DOI: 10.24425/jwld.2023.147242

Author Response

Comments 1: Your article makes poor reference to heat and power consumption. Table 1 contains the following:

1. Energy Management: What influences Overall energy usage?

What constitutes overall energy usage?

Compare in Literature Review with buildings in middle climates [x1].

[x1] Kapalo, P.; Adamski, M. The Analysis of Heat Consumption in the Selected City. Lecture Notes in Civil Engineering, 2021. 100 LNCE pp. 158 - 165

DOI: 10.1007/978-3-030-57340-9_20

Response 1: We sincerely thank the reviewer for this valuable comment. We fully agree that it is important to clarify what constitutes the overall energy usage analyzed in this research. Accordingly, we have added further explanation and integrated a methodology to specify the components of total energy consumption, as follows:

“Previous studies highlight that overall energy consumption in campus buildings typically includes electrical loads (lighting, plug loads, and HVAC systems) and thermal energy for heating or hot water supply, depending on the climatic context. For example, Kapalo and Adamski [x1] analyzed heat consumption in buildings located in moderate climates, where heating demand strongly influenced overall usage. In contrast, in tropical climates such as Thailand, electricity dominates total energy consumption, with HVAC and lighting being the primary contributors.” – page 5, paragraph 4, and line 212-218.

And incorporated into the discussion the point that “In our case study, “overall energy usage” refers specifically to electricity consumption across 25 faculty buildings, as heating loads are negligible in this climatic region. Stakeholders prioritized energy management because electricity accounts for the largest share of operating costs, mainly driven by HVAC systems and lighting demand.” – page 14, paragraph 2, and line 456-459.

Comments 2: 5. Water Flow Tracking. Does this refer to hot water, cold water, or total water consumption?

The issue of pipeline anomalies is also worthy of consideration [x2], with regard to their types, causes and effects.

[x2] Kalenik, M.; Chalecki, M.; Wichowski, P.; Kiczko, A.; Chmielowski, K.; Świętochowska, M.; Gwoździej-Mazur J. Real values of local resistance coefficients during water flow through a pipe aerator with filling, Journal of Water and Land Development, 2023 59, pp. 174 - 182

DOI: 10.24425/jwld.2023.147242

Response 2: We sincerely thank the reviewer for this valuable comment. We agree with adding information on water flow detection, as the water used inside and outside the building consists of multiple types. Therefore, I will specify that:

“Stakeholders defined the water flow tracking function to focus primarily on cold and total water consumption, rather than hot water, as heating is not a significant demand in the local context. One challenge identified was the difficulty of fully modeling and monitoring the entire piping network within buildings, which would require substantial resources. Instead, a modular approach was proposed, allowing future expansion to cover more detailed monitoring. Furthermore, pipeline anomalies (e.g., leaks, blockages, and abnormal resistance) were considered relevant to water management, consistent with findings in [x2], which emphasize the role of local resistance coefficients in flow behavior and anomaly detection.” – page 5-6, paragraph 5, 1, and line 219-227.

Comments 3: What are the shares of functional requirements identified by stakeholders:

In total energy consumption?

In total operating costs?

Response 3: Thank you for your comment. We agree to explain the proportion of operational requirements specified by stakeholders, including energy use and total operational costs. We have provided further details in discussion.

“Stakeholders ranked energy management as the highest-priority functional requirement, reflecting its dominant contribution to total operating costs. Although precise cost breakdowns vary by building, electricity expenditures typically account for 60–70% of operating budgets in campus facilities, consistent with values reported in the literature. The second-ranked requirement, maintenance alerts, was associated with extending equipment life cycles and reducing unexpected repair costs. At the same time, water flow tracking, although less prioritized, was recognized as a potential area for long-term sustainability planning.” – page 14, paragraph 5, and line 469-476.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The revised manuscript “A Dynamic Digital Twin Framework for Sustainable Facility Management in a Smart Campus: A Case Study of Chiang Mai University” shows clear improvement and demonstrates that the authors have carefully addressed most of the earlier comments and integrated the suggested revisions. The methodology section has been significantly strengthened with additional clarity on the framework design, implementation process, and validation approach, which enhances the overall rigor and transparency of the study. The inclusion of more detailed case study insights from Chiang Mai University adds practical depth and makes the contribution more tangible. The expanded discussion on sustainability indicators and their integration within the digital twin framework provides a stronger connection between theory and application. Furthermore, the manuscript now incorporates relevant literature and international perspectives, which improves the contextualization of the research within broader smart campus and facility management discourses. Overall, the revised manuscript has responded effectively to reviewer feedback, substantially enhancing both clarity and scholarly value. I recommend acceptance after minor revisions, focusing on refining the conclusion, improving visual clarity of framework representations, and adding a short note on limitations and future research directions.

Author Response

Comments 1: The revised manuscript “A Dynamic Digital Twin Framework for Sustainable Facility Management in a Smart Campus: A Case Study of Chiang Mai University” shows clear improvement and demonstrates that the authors have carefully addressed most of the earlier comments and integrated the suggested revisions. The methodology section has been significantly strengthened with additional clarity on the framework design, implementation process, and validation approach, which enhances the overall rigor and transparency of the study. The inclusion of more detailed case study insights from Chiang Mai University adds practical depth and makes the contribution more tangible. The expanded discussion on sustainability indicators and their integration within the digital twin framework provides a stronger connection between theory and application. Furthermore, the manuscript now incorporates relevant literature and international perspectives, which improves the contextualization of the research within broader smart campus and facility management discourses. Overall, the revised manuscript has responded effectively to reviewer feedback, substantially enhancing both clarity and scholarly value. I recommend acceptance after minor revisions, focusing on refining the conclusion, improving visual clarity of framework representations, and adding a short note on limitations and future research directions.

Response 1: Thank you for your comment and we have revised the Conclusion section to make it more concise and to highlight the key contributions more clearly. In addition, we have added a short note on the limitations and future research directions, as suggested. Regarding the framework figures, we have adjusted the image size and enlarged the text that was previously too small to improve readability.

“Improve conclusion:

This study demonstrates the successful design, development, and deployment of a digital twin framework for sustainable facility management at Chiang Mai University. The system transforms fragmented information into an interactive decision-support tool by integrating 3D spatial modeling, real-time IoT sensor data, and multiscale dashboard visualization. The case study highlights its modular and replicable architecture, validation through real-world implementation, and contribution to aligning digital twin applications with sustainability objectives. Beyond campus operations, the framework offers broader implications for data-driven and climate-responsive smart city governance.

One of the system's primary strengths is its ability to synthesize complex and heterogeneous datasets into intuitive, user-centric visualizations. The multilevel dashboard allows stakeholders to monitor aggregate energy consumption, identify anomalies, and assess space utilization dynamically. This functionality supports operational efficiency and targeted interventions—such as detecting underutilized spaces or identifying devices left in standby mode.

Notably, user feedback highlighted the value of exportable energy data, which extends the platform's utility into downstream workflows such as monthly reporting, performance benchmarking, and machine learning–based forecasting. The system's modular and scalable architecture, built on open-source tools (e.g., Firebase, QGIS, SuperMap iDesktop, and web-based dashboards) and standardized APIs, further ensures flexibility for future expansion. Core modules, including data ingestion, validation, and API interfacing, are hardware-agnostic, allowing seamless integration of additional sensors and third-party data services as infrastructure evolves. Open platforms reduce costs, enhance interoperability, and ensure replicability across institutional and urban contexts.

The visual interface is critical in translating complex sensor data into actionable insights. At the faculty level, administrators can compare energy use across buildings. In contrast, at the room level, real-time monitoring of lighting and HVAC operations enables detection of inefficiencies or system faults. These layered dashboards enhance transparency, enable proactive management, and foster sustainable energy practices.

Nonetheless, the current system has limitations. Despite these contributions, some functions—such as equipment data entry—still require manual input, and expanding the framework to include occupancy tracking, water monitoring, or automated control will add complexity. Future research and development will therefore focus on addressing these challenges by prioritizing the automation of data collection through advanced sensing and edge computing, enhancing interoperability via open data standards, and strengthening data governance and cybersecurity frameworks. Furthermore, incorporating predictive analytics and machine learning will enable the transition from a monitoring-oriented platform to a proactive optimization system capable of generating recommendations and supporting informed interventions.

Beyond the campus context, the framework's modular and hardware-agnostic architecture provides strong potential for replication in smart city environments. Extending its functions to urban-scale domains—such as energy districts, water distribution networks, and mobility systems—positions the platform to contribute more broadly to sustainability objectives, including carbon reduction, resource efficiency, and climate resilience.

” – Conclusion, page 24, paragraph 1-6, and line 721-763.

– Figure 3-8, page 10-13.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Fig. 16:

What means:

FLOOR 08 ?

ROOM 18 ?

-------

kW/hour or kWh (SI units) or other meaning?

Air ?

Light ?

Socket ?

--------

Fig. 17:

For PM 2.5 max is between min and mean values in Mar. - July?

Max values of temperature are near 0^oC?

Max values of humidity take 0?

The right part of the figure is given twice.

Could you complete this drawing with temperature, humidity and dust data for years 2023- 2024?

Author Response

Comments 1:

Fig. 16:

What means:

FLOOR 08?

ROOM 18?

kW/hour or kWh (SI units) or other meaning?

Air, Light, and Socket?

Response 1: Thank you for your comment and we would like to clarify that the labels shown are identifiers in the 3D model database. For example, “Floor 08” refers to the 8th floor of the 30-Year Engineering Building, while “Room 18” denotes the internal ID for a specific room, which in this case corresponds to Room 818: the City Research and Development Center.

In addition, the reported value refers to energy consumption. In accordance with SI units, it should be expressed as kWh (kilowatt-hours), not kW/hour. The system employs PZEM-004T power sensors, which measure both instantaneous power (kW) and accumulated energy (kWh). The error occurred during the dashboard development stage and has now been corrected in the revised version.

For clarity, the category labels in the dashboard represent the following:

Air = air-conditioning unit(s)

Light = lighting system

Socket = plug loads (e.g., computers, appliances, chargers)

These clarifications and corrections have been incorporated into the revised manuscript to improve the accuracy and readability of Figure 16.

“

The room-level dashboard provides detailed information for individual rooms. On the left, it displays environmental conditions, while the right side shows the room’s total energy consumption. In this example, identifiers such as FLOOR 08 and ROOM 18 correspond to the 8th floor of the 30-Year Engineering Building and a specific room (Room 818: the City Research and Development Center) in the 3D model database. Energy usage is measured in kilowatt-hours (kWh), the correct SI unit, using sensors (PZEM-004T) that capture both instantaneous power (kW) and accumulated energy (kWh). Consumption is further disaggregated into three categories—Air (air-conditioning units), Light (lighting system), and Socket (plug loads such as computers and appliances)—which are represented in bar chart form. This breakdown enables users to assess the relative impact of different types of electrical appliances within the room. Additionally, the dashboard includes a dedicated function for maintenance technicians, allowing them to manage and update appliance data in the database. This feature enhances data completeness and supports the potential development of predictive maintenance and repair alert systems in the future.

” – page 21, paragraph 1, and line 594-608.

Comments 2:

For PM 2.5 max is between min and mean values in Mar. - July?

Max values of temperature are near 0^oC?

Max values of humidity take 0?

The right part of the figure is given twice.

Response 2: We appreciate the reviewer’s observation on Figure 17. We acknowledge that this was indeed an error. After carefully reviewing both the database and the computational processes, we identified that the issue most likely originated from a programming error. Corrective actions have been taken, and the system has been improved to ensure that the graphs are now rendered accurately in the revised version. – page 20, figure 16-17.

Comments 3: Could you complete this drawing with temperature, humidity and dust data for years 2023- 2024?

Response 3: We thank the reviewer for the valuable comment. As the dashboard was originally programmed to display temperature, humidity, and dust data with a retrospective window of six months, extending it to cover the years 2023–2024 would require substantial modification of the underlying system, including adjustments to the user interface, revisions to data transmission protocols, and restructuring of the database architecture. Therefore, such an extension is not straightforward within the current framework.

Nevertheless, we have corrected the graph as presented in Figure 17. The data now cover the retrospective period from July to August 2024, ensuring consistency with the situations discussed and alignment with the overall analysis of the 2023–2024 dataset. – page 20, figure 17.