Data Center Developments for Flexible Generation Dispatch, Advanced Infrastructure, and Ultra-Fast Digital Twins

Intelligent Transportation Systems (ITS) have progressively incorporated machine learning to optimize traffic efficiency, enhance safety, and improve real-time decision-making. However, the traditional centralized machine learning (ML) paradigm faces critical limitations regarding data privacy, scalability, and single-point vulnerabilities. This study explores FL as a decentralized alternative that preserves privacy by training local models without transferring raw data. Based on a systematic literature review encompassing 39 ITS-related studies, this work classifies applications according to their architectural detail—distinguishing systems from models—and identifies three families of federated learning (FL) frameworks: privacy-focused, integrable, and advanced infrastructure. Three representative frameworks—Federated Learning-based Gated Recurrent Unit (FedGRU), Digital Twin + Hierarchical Federated Learning (DT + HFL), and Transfer Learning with Convolutional Neural Networks (TFL-CNN)—were comparatively analyzed against a client–server baseline to assess their suitability for ITS adaptation. Our qualitative, architecture-level comparison suggests that DT + HFL and TFL-CNN, characterized by hierarchical aggregation and edge-level coordination, are conceptually better aligned with scalability and stability requirements in vehicular and traffic deployments than pure client–server baselines. FedGRU, while conceptually relevant as a meta-framework for coordinating multiple organizational models, is primarily intended as a complementary reference rather than as a standalone architecture for large-scale ITS deployment. Through application-level evaluations—including traffic prediction, accident detection, transport-mode identification, and driver profiling—this study demonstrates that FL can be effectively integrated into ITS with moderate architectural adjustments. This work does not introduce new experimental results; instead, it provides a qualitative, architecture-level comparison and adaptation guideline to support the migration of ITS applications toward federated learning. Overall, the results establish a solid methodological foundation for migrating centralized ITS architectures toward federated, privacy-preserving intelligence, in alignment with the evolution of edge and 6G infrastructures.

The energy domain worldwide is experiencing high transformative pressure due to the imperative of climate change and new opportunities brought by various rapidly evolving digital technologies. Particularly, this change is affecting smart grids (SGs), which are increasingly shifting toward utilizing renewable and distributed energy sources. The natural intermittency of these energy sources increases the complexity of SG operations, such as ensuring continuity of energy supply and demand response balancing. In mitigation of these challenges, the tools and complex approaches that digitalization can provide have shown themselves particularly advantageous. There is a solid body of work showing how technologies like the Internet of Things (IoT), distributed ledgers, edge and cloud computing, machine learning (ML), etc., can be applied to address a variety of technical and economic problems in the energy sector, emphasizing SGs. This paper presents the most comprehensive literature review to date on digitalization in renewable energy source-based SGs, synthesizing over 200 studies across data analytics, artificial intelligence and ML, digital twins, edge–fog–cloud computing, the IoT, advanced metering infrastructure, and distributed ledger technologies. Unlike previous reviews, which are often limited to a single technology or narrow application, this work provides a cross-technology synthesis linking technical and financial aspects, identifies consolidated research gaps, and proposes a unified research agenda. The review further highlights future trends, including large language models, 6G communications, and distributed autonomous organizations, and discusses their implications for both industry practice and academic research.

Data centres in contemporary times are essential as the supply of data increases. Data centres are areas where computing systems are concentrated for facilitating data processing, transfer and storage. At present, traditional data centres have moved more towards the cloud model—thereby making the processing, storage and harnessing of data more manageable and more accessible via the utility and subscription-based model of computing services. From the administrative point of view, cloud data centres are complex systems and hard to grasp and require large amounts of time to analyse different aspects of the cloud data centre such as maintenance and resource management. For a cloud data centre admin, this could be a challenging problem and a highly time-consuming task. Accordingly, there is a need to improve the useability of cloud data centre monitoring and management tools, and the digital twin could fulfil this need. This book chapter’s primary objective is to construct a digital twin—a 3D visualisation and monitoring tool—of a cloud data centre managed by OpenStack, the well-known open-source cloud computing infrastructure software. To evaluate our proposed tool, we garner feedback on the digital twin’s useability compared to the OpenStack dashboard. The input will be received from cloud data centre experts as they test the digital twin and answer various questions in an interview. The study results show that our proposed digital twin will help data centre admins better monitor and manage their data centres. It will also facilitate further research and implementation of the digital twin of data centres to improve usability.KeywordsCloud data centreDigital twinOpenStackVisualisationMonitoringResource management

Real-Time Machine Learning-based fault Detection, Classification, and locating in large scale solar Energy-Based Systems: Digital twin simulation

Collaboration is a cornerstone of digital transformation in the Architecture, Engineering, and Construction (AEC) sector. Despite widespread adoption of building information modeling (BIM), a critical “human factor gap”—the disconnect between rapid technological advancement and organizational readiness—remains largely unaddressed. This study investigates the intellectual structure of BIM collaboration research. Adopting a mixed-methods review design, it conducts a comprehensive bibliometric analysis of 1595 journal articles indexed in Web of Science and Scopus, followed by a qualitative content analysis of 38 purposefully selected studies. Findings reveal a clear shift from process-oriented collaboration frameworks toward data-centric and cyber–physical systems, including digital twins, Artificial Intelligence, and blockchain-enabled environments. However, this evolution also highlights a notable asymmetry in BIM collaboration, with technology-driven dynamics substantially outweighing human and organizational considerations. This imbalance, conceptualized as a human factor gap in thematic development observed through the bibliometric analysis, points to a potential socio-technical misalignment, in which technological topics receive more scholarly attention than organizational and human factors. Rather than framing collaboration maturity as a purely technical challenge, the study emphasizes the need for advanced digital infrastructure, human-centered governance models, organizational readiness, and socio-technical competencies. It provides analytical insights for researchers and implications for policymakers and industry stakeholders, emphasizing the necessity of integrated approaches that address both the technological and human dimensions of BIM-enabled collaboration.

Industrial automation is expected to benefit from the capabilities of Beyond 5G (B5G)/6G, particularly in improving real-time decision making and control in information-driven scenarios, supporting improved efficiency, precision and safety in connected manufacturing environments. While not all aspects of industrial automation are equally affected, certain areas such as real-time monitoring, wireless control and predictive maintenance can benefit significantly from the high reliability, ultra-low latency and scalability of next-generation wireless systems. Motivated by the increasing convergence between operational technology (OT) and advanced communication infrastructures, this paper surveys how B5G/6G can serve as an enabler for data-centric industrial ecosystems. The main contribution lies in analysing evolving industrial architectures, the integration of OT systems and application domains including digital twins, cloud robotics and smart devices. At the end of the paper, we also offer strategic insights into enabling technologies and future challenges, providing a comprehensive overview of how B5G/6G can support the evolution of Industry 5.0.

This study proposes the development of an AI-driven digital twin for data centers aimed at improving energy efficiency, reducing carbon footprint, and enhancing operational performance. The digital twin—a virtual replica of the physical data center—will be equipped with real-time AI algorithms to predict thermal loads, analyze cooling requirements, and automatically adjust operations to minimize energy consumption. This paper explores the integration of AI with digital twin architectures, tests its performance in simulated scenarios, and evaluates potential energy savings as well as contributions to Green IT practices.

Autonomous digital twin framework for gas turbine combined cycle control loops: Comparative study of proportional-integral control, reinforcement learning, and reinforcement learning with agents

Digital twins provide living digital models of physical systems that enable data-driven analysis and application of artificial intelligence to better manage selective aspects of the data center and drive efficiency for sustainability.

SeaDataNet is a pan-European infrastructure that manages and provides access to marine datasets collected by European organisations through research cruises and observational activities in coastal waters, regional seas, and the global ocean. It was founded by National Oceanographic Data Centres (NODCs) and major marine research institutes. The network has expanded through successive EU-funded RTD projects and by contributing to major European initiatives such as EMODnet, Copernicus Marine Service, ENVRI, and the European Open Science Cloud (EOSC).SeaDataNet develops and promotes widely adopted standards, vocabularies, software tools, and services that support FAIR marine data management. Its core service, the CDI (Common Data Index), provides unified online discovery and access to in situ marine observation data managed by more than 115 data centres in 34 countries. The service currently offers access to over 3 million datasets from more than 1,000 European organisations, covering physical, chemical, biological, geological, and geophysical data from European waters and the global ocean. The use of standard metadata, formats, and controlled vocabularies ensures rich, highly FAIR datasets.SeaDataNet also delivers core data services for EMODnet Chemistry, Bathymetry, and Physics, harmonising large volumes of marine data that support the production of thematic data products, including Essential Climate Variable (ECV) and Essential Ocean Variable (EOV) datasets.Environmental science increasingly relies on large, heterogeneous, and rapidly growing data collections that must be efficiently accessed, subsetted, and harmonised for use in models, digital twins, AI workflows, and Virtual Research Environments (VREs). The fully open-source Beacon software, developed by MARIS https://beacon.maris.nl/, addresses these challenges by enabling cloud-native, high-performance data lakes that are fast to deploy and access. Beacon supports parameter harmonisation using metadata annotations based on NERC vocabularies, ECV vocabularies, and the I-ADOPT methodology adopted in ENVRI-HUB Next.To improve the ease of access to subsets of the SeaDataNet CDI data collection, a Beacon instance containing all the open SeaDataNet data was set-up. This now allows users to obtain real-time access to data subsets in multiple data formats (NetCDF, Parquet, Zarr) and flexible querying from Jupyter Notebooks or a newly developed Beacon studio (user interface) for non-technical users. Within ENVRI-HUB Next, this SeaDataNet instance enables on-the-fly access to ECV-related subsets from millions of files via Jupyter Notebooks, ready for use in the Analytical Framework.The presentation focuses on this use case, the technical solution, and its potential applicability for other Research Infrastructures supporting EOSC use cases.

Facing the high power and heat density of the AI requirement, this paper reviews recent advances in multi-level thermal performance assessment for data centers. This multi-level framework covering chip, device, rack/row, and room levels. Corresponding to the multi-level framework, a metric system micro-scale and macro-scale indicators is developed to improve system resilience, facility safety and overall energy efficiency. Further, this review summarize progress in data collection, including advanced sensor technologies, computational fluid dynamics (CFD) and reduced-order models, such as AI/ML-based digital twins. Based on recent advances, future research directions are discussed in this paper, including standardization of multi-level metrics, high-accuracy data collection technologies, and utilization of AI/ML based technologies.

Background: Data centers are major energy consumers, with cooling systems accounting for a significant portion of operational costs and carbon emissions. Recent advances in artificial intelligence (AI) offer promising strategies for optimizing thermal management. However, a comprehensive overview of publication trends, thematic evolution, and collaborative networks in AI-enabled data-center cooling is lacking. Methods: We conducted a bibliometric analysis of 311 peer-reviewed articles retrieved from the Web of Science database, covering the period 2015–2024. Using VOSviewer and CiteSpace, we mapped annual publication trends, country-level contributions, co-authorship networks, co-citation structures, and keyword co-occurrence clusters. Burst-detection analysis identified emerging topics and their temporal dynamics. Results: Annual publications grew from 2 in 2015 to 80 in 2024, reflecting accelerating research interest. China (28 %) and the United States (25 %) led author-country instances, with strong bilateral collaborations. Co-occurrence analysis revealed 12 thematic clusters (modularity Q = 0.72, silhouette = 0.91), the largest of which focused on “reinforcement learning,” followed by “energy efficiency” and “free cooling.” The ten most frequent keywords—e.g., “data center cooling,” “machine learning,” “thermal management”—underscored the dominance of AI-driven control strategies. Burst analysis showed an early focus (2015–2018) on “neural network” and “genetic algorithm,” transitioning (2019–2024) to “reinforcement learning,” “deep reinforcement learning,” and, more recently (2022–2024), “digital twin” and “meta-learning.” Discussion: The field has matured from initial explorations of classical machine-learning algorithms to advanced, adaptive reinforcement-learning frameworks and simulation-based optimization. Real-world implementations—such as RL-based controllers deployed by major technology firms—demonstrate tangible energy savings. Future research should pursue multiobjective optimization, enhance digital-twin fidelity to bridge the sim-to-real gap, develop interpretable AI policies, and incorporate non-English literature and industry reports to broaden the knowledge base.

The article examines the impact of artificial intelligence (AI) on the transformation of the educational and scientific spheres. It is determined that today, AI is evolving from a simple technology into a powerful tool for transforming education and science. The use of AI offers several advantages, including personalization of the learning process through adaptive systems, automation of routine tasks, and the utilisation of data analytics tools to integrate research and data analysis into scientific and educational cycles, such as those facilitated by intelligent laboratories and digital twins. At the same time, AI can pose risks in areas such as academic integrity, research ethics, publications, and peer review; traditional assessment and teaching methods become irrelevant and ineffective. The use of AI in educational and scientific activities not only improves the teaching process, but also brings universities closer to achieving sustainable development goals, ensuring a higher level of educational services through the use of digital tools, adaptive technologies, learning stimulators, increasing the inclusiveness of the educational environment, increasing the energy efficiency of higher education institutions, training specialists in the development and use of AI for Industry x.0. At the same time, this process negatively affects SDG 7 and 13 due to the increase in electricity consumption by data centres and the carbon footprint of supercomputers. Analysis of international practices reveals that leading universities prioritise innovation in the academic environment by leveraging AI tools. A university's innovative AI-oriented ecosystem is being formed, which combines educational and research initiatives through ethical principles, transparency, a focus on interdisciplinary approaches, and the development of cooperation. Focusing on the market, universities must also adapt their educational products, taking into account requirements such as skill transformation, automation of routine roles, and the demand for AI specialists. It is shown that the educational services market itself is transforming in response to the demands of the time, including the formation of data analytics, machine learning, and programming competencies, the development of skills in using AI tools in the workplace, and a deepening of knowledge about advanced developments in the field of AI. The following areas of educational services are highlighted: university educational programs, courses (for beginners or specialised, such as those on studying Power BI), comprehensive boot camps, intensive training, and self-study. The price range, depending on the profile and volume of such educational services in Ukraine, was analysed, which ranges from 11,000 UAH for short-term courses to 332,800 UAH per year for bachelor's programs in the speciality «Systems Analysis and Data Science». It is demonstrated that enhancing the level of AI competencies is a promising approach to increasing a specialist's competitiveness in the labour market. For the education and science system, it is crucial to select training forms and methods that most effectively meet the current requirements.

Many data centers, medical complexes, government facilities, factories and other commercial buildings have standby electrical generators. Quiet for most of the year, the generators are only used in the event of a power outage or for periodic testing. Portland General Electric (PGE) offers a program aimed at supplying capacity resources that utilizes these hidden resources to their full potential as a single aggregated generating peaking plant. PGE's "dispatchable standby generation" (DSG) program currently ties together 15 MWs of standby generators to support peak power demand on PGE's system and to ensure reliable service for customers. This power can also help avoid buying wholesale power when prices are skyrocketing or may allow less expensive resources to be utilized while DSG power can be counted as reserve capacity. DSG generator power first supplies power to its designed load at the facility and any excess power flows into the PGE system. To the PGE system grid, this appears as a drop in load and a small increase in supply. Finally, PGE links the GenOnSys application to PGE's system control center's energy management system via a secure fiber connection.

Backup flexibility classes in emerging large-scale renewable electricity systems