The mission of the Big Data Engineering, Analytics & Science Research Group is to perform applied state of the art research on the R&D domains of Big Data Engineering and Big Data Science. The DEA Research Group aims at pushing the boundaries of large-scale data systems and big-data driven intelligent analytics to solve real-world challenges across various application domains including e-Health, FinTech, Agriculture, and more.
Big Data Engineering: The DEA Research Group undertakes the design, implementation and deployment of high-performance, scalable and efficient big data architectures of data-driven platforms and systems, leveraging open-source distributed frameworks facilitating seamless (real-time and non-real-time) data ingestion, data cleaning, data anonymization, data harmonization, and data processing.
Big Data Science: The DEA Research Group undertakes the design, implementation and deployment of accurate, robust and trustworthy ML/DL-powered workflows leveraging open-source M/L and D/L libraries and frameworks, optimizing AI-empowered models for predictive analytics, NLP, and computer vision.
Key Research Areas
-
Big Data Warehousing: Our research in Big Data Warehousing focuses on designing scalable, high-performance architectures capable of efficiently storing, managing, and querying massive datasets across distributed environments. We explore next-generation data architectures, including Data Lakes, Data Fabrics and Data Meshes, leveraging open-source technologies like Apache Hive, Apache Iceberg and more. We investigate automated metadata management, query optimization, and federated querying across heterogeneous data sources, while we also research privacy-aware warehousing.
-
User Authentication & Fine-Grained Access Control: Our research in user authentication and fine-grained access control focuses on developing secure, adaptive, and privacy-preserving mechanisms for modern distributed systems. We investigate various access control models, including attribute-based (ABAC) and role-based (RBAC) access control models enhanced with machine learning for dynamic policy enforcement. We also study blockchain-based decentralized identity solutions (e.g., SSI, DID) for privacy-preserving authorization.
-
Streaming Data Handling: Our research in streaming data handling focuses on building real-time, scalable, and fault-tolerant architectures for high-velocity data processing. We investigate distributed stream processing frameworks (e.g., Apache Flink, Kafka Streams, Spark Structured Streaming) to enable low-latency event processing, complex event detection, and stateful stream analytics, while we also explore streaming machine learning for online model training and inference.
-
Synthetic Data Generation: Our research in synthetic data generation focuses on developing privacy-preserving, high-fidelity data replicas that maintain statistical validity while eliminating exposure risks. We explore deep generative models (e.g., GANs, diffusion models) and differential privacy techniques to produce realistic yet anonymized datasets for sensitive domains like healthcare and finance. We also investigate federated synthetic data creation in distributed environments, as well as synthetic data for bias mitigation, enhancing model robustness.
-
Multi-Modal Data Fusion: Our research in multi-modal data fusion focuses on developing advanced methodologies to integrate and analyze heterogeneous data types, including text, images, sensor signals, and structured data, for richer, more robust AI systems. We investigate both standard machine learning and deep learning architectures (e.g., cross-modal transformers, graph neural networks) to model complex interactions between modalities. A key focus is explainable fusion models to enhance interpretability in critical domains like healthcare.
-
Explainability & Interpretability: Our research in eXplainable AI (XAI) and Mechanistic Interpretability (Mech. Interp.) focuses on understanding and explaining the behaviour of complex AI systems. We investigate XAI post-hoc explanation methods (e.g., SHAP, LIME) and intrinsically interpretable models (e.g., decision trees, rule-based systems) to uncover the reasoning behind complex AI decisions and produce human-readable insights. With advanced Mech. Interp. Methods we further explore the inner model workings (e.g. activation visualization, neuron attribution, TransformerLens). A key A key emphasis of our work is emphasis is on domain-specific explanations, tailored to stakeholders in healthcare, finance, and other high-stakes fields. We also research fairness-aware XAI, detecting and mitigating biases in black-box models.
-
Meta-Learning: Our research in meta-learning focuses on developing adaptive AI systems that can rapidly acquire new skills or generalize across tasks with minimal data requirements and supervision. We investigate optimization-based approaches that learn effective parameter initializations for fast fine-tuning, as well as metric-based methods for few-shot, low-shot problems. We also explore Bayesian meta-learning to quantify uncertainty and cross-domain meta-generalization, as well as automated meta-learning pipelines that dynamically adjust learning strategies based on task characteristics.
The Big Data Engineering, Analytics & Science Research Group focuses on two major R&D domains: Big Data Engineering and Big Data Science.
Big Data Engineering
Our group focuses on foundational and emerging areas in Data Engineering, with a strong emphasis on scalable architecture, real-time data processing, robust data governance, and secure, self-serve analytics infrastructure. Key technology domains include:
-
Modern Data Stack & Cloud-Native Architectures: We engineer modular, cloud-native data platforms designed for scalability, resilience, and operational efficiency. Our architecture supports both batch and streaming workloads, with a focus on decoupling compute and storage, enabling elastic resource allocation and cost optimization across the data lifecycle. Key design principles include declarative infrastructure, composable pipelines, and extensible tooling, allowing for rapid development, reproducibility, and infrastructure-as-code practices. This architecture supports hybrid and multi-cloud strategies, while enabling data teams to scale operations without compromising maintainability or performance. Technologies: Apache Iceberg, MinIO, Trino, Kubernetes, Apache Airflow
-
Identity & Access Management (IAM) for Data Platforms: We build secure, flexible identity and access management systems designed to meet the complex needs of modern data ecosystems. Our approach is centered around Attribute-Based Access Control (ABAC), enabling dynamic, fine-grained permissions based on user attributes, data sensitivity, and operational context. Technologies: Keycloak, OpenID Connect (OIDC), OAuth2, ABAC, RBAC,
-
Open Policy Agent (OPA) Streaming & Real-Time Data Infrastructures: Designing end-to-end event-driven systems for low-latency data processing. Focus includes event sourcing, CDC (Change Data Capture), and scalable stream enrichment and aggregation for real-time analytics and alerting. Technologies: Apache Kafka, Apache Flink, Debezium.
Specialized Expertise
Our team brings deep expertise across critical subdomains of data engineering, enabling the development of advanced, production-grade data systems. We specialize in the design and implementation of complex data workflows, high-throughput ingestion pipelines, and robust metadata and lineage frameworks. Our work spans both foundational and cutting-edge techniques that bridge infrastructure, architecture, and applied data engineering. Key focus areas include:
-
Cost-Efficient Data Infrastructure: Designing systems with workload-aware optimization strategies such as storage tiering, query pushdown, compute auto-scaling, and smart caching to minimize cost without sacrificing performance.
-
Hybrid & Multi-Cloud Strategy: Architecting vendor-neutral platforms that support workload portability and interoperability across cloud providers, enabling resilience and avoiding vendor lock-in.
-
Secure, Policy-Driven Data Access: Implementing layered, policy-as-code access controls with dynamic evaluation based on user attributes and data context, ensuring both regulatory compliance and developer autonomy.
Big Data Science
Our group focuses on foundational and emerging areas in AI, with a strong emphasis on adaptive intelligence, trustworthiness, and cross-modal data integration. Key technology domains include:
-
Synthetic Data Generation & Privacy-Preserving AI: Developing high-fidelity, privacy-preserving synthetic datasets using deep generative models (e.g., GANs, diffusion models) and differential privacy. Work includes federated synthetic data generation and bias mitigation techniques for sensitive domains such as healthcare and finance. Technologies: PyTorch, TensorFlow, Opacus, SynthCity.
-
Multi-Modal Data Fusion: Creating robust AI systems through the integration of heterogeneous data types including text, images, time-series, and structured data. Emphasis is placed on architectures such as cross-modal transformers and graph neural networks, with explainability in critical applications. Technologies: HuggingFace, Transformers, PyTorch Geometric, DGL, CLIP.
-
Explainability & Mechanistic Interpretability: Advancing both post-hoc explainability (e.g., SHAP, LIME) and mechanistic interpretability by probing model internals via techniques like activation visualization and neuron attribution. Applied in fairness-sensitive contexts to build trust in high-stakes decisions. Tools and frameworks: Captum, TransformerLens, OpenAI Microscope, SHAP, LIME.
-
Meta-Learning & Few-Shot Adaptation: Designing adaptive AI systems that learn efficiently from limited data. Research spans optimisation-based meta-learning (e.g., MAML), metric-based methods, Bayesian approaches, and automated pipelines for task- specific strategy tuning. Technologies: learn2learn, higher (PyTorch), Meta-World, Pyro.
Specialised Expertise
Our group’s core competencies lie in designing AI systems that are high-performing, interpretable, adaptive, and privacy-conscious. Our specialised areas of expertise include:-
Mechanistic Interpretability of Deep Networks: We go beyond surface-level explainability by mapping and analysing the internal representations and decision pathways of neural models, using frameworks like TransformerLens. This enables grounded model understanding, especially in mission-critical domains.
-
Few-Shot and Cross-Domain Generalisation: Our work in meta-learning allows us to create models capable of adapting to new environments with minimal data, which is vital for real-world deployment in dynamic settings. We specialise in optimisation-initialisation strategies and Bayesian uncertainty modelling.
-
Federated and Differentially Private Synthetic Data: Addressing data scarcity and privacy simultaneously, we develop synthetic data pipelines suitable for decentralized training environments without compromising utility or safety.
-
Multi-Modal Reasoning and Fusion Models: We have deep expertise in integrating diverse data modalities using advanced neural architectures, enabling richer representations and improved performance in complex tasks. These include applications in healthcare diagnostics, financial risk modelling, and autonomous sensing systems.
-
Fairness-Aware AI Systems: Our research emphasises bias detection and mitigation in both traditional and deep learning systems, for equitable outcomes across diverse user populations. Our techniques are applied in compliance-sensitive sectors including amongst others healthcare, and finance.
Description
AI-DAPT brings forward a data-centric mentality in AI, that is effectively fused with a model-centric, science-guided approach, across the complete lifecycle of AI-Ops, by introducing end-to-end automation and AI-based systematic methods to support the design, the execution, the observability and the lifecycle management of robust, intelligent and scalable data-AI pipelines that continuously learn and adapt based on their context. AI-DAPT will design a novel AI-Ops / intelligent pipeline lifecycle framework cross-cutting the different business, legal/ethics, data, AI logic/models, and system requirements while always ensuring a human-in-the-loop (HITL) approach across five axis: “Data Design for AI”, “Data Nurturing for AI”, “Data Generation for AI”, “Model Delivery for AI”, “Data-Model Optimization for AI”. AI-DAPT will contribute to the current research and advance the state-of-the-art techniques and technologies across a number of research paths, including sophisticated Explainable AI (XAI)-driven data operations from purposing, harvesting/mining, exploration, documentation and valuation to interoperability, annotation, cleaning, augmentation and bias detection; collaborative feature engineering minimizing the data where appropriate; adaptive AI for model retraining purposes. Overall, AI-DAPT aims at reinstating the pure data-related work in its rightful place in AI and at reinforcing the generalizability, reliability, trustworthiness and fairness of Al solutionsKey Contributions
Within AI-DAPT, UBITECH will design the AI-DAPT reference architecture and produce design guidelines that showcase how the AI-DAPT results can be integrated and adopted in different, existing AI platforms; design the APIs of AI-DAPT services; elaborate an integration plan which will govern the integration and ML-Ops activities (code maintenance, continuous integration, software evaluation); provide the backbone execution environment; integrate the services of Reference Architecture to deliver the AI-DAPT Platform; verify and validate the proper operation of AI-DAPT Platform and take corrective actions if any defects are encountered; investigate and perform integration of AI-DAPT Framework to existing related AI solutions. Moreover, UBITECH will significantly contribute to the implementation of the Data Curation Methods & Services, the Data Generation Methods & Services, as well as the Hybrid Science-AI Model Interactive Training mechanisms within AI Pipelines.
Description
AUTO-TWIN addresses the technological shortcoming and economic liability of the current systemengineering model by 1) introducing a breakthrough method for automated process-aware discovery towards autonomous Digital Twins generation, to support trustworthy business processes in circular economies; 2) adopting an (International Data Space) IDSbased common data space, to promote and facilitate the secure and seamless exchange of manufacturing/product/business data within value-networks in a circular-economy ecosystem; 3) integrating novel hardware technologies into the digital thread, to create smart Green Gateways, empowering companies to perform data and digital twin enabled green decisions, and to unleash their full potential for actual zero-waste Circular Economy and reduced dependency from raw materialsKey Contributions
Within AUTO-TWIN, UBITECH implements a decentralized blockchain-based approach to secure and guarantee key data and information in the Twin lifecycle. Technologically, the blockchain implementation will be based on the Hyperledger Fabric and will leverage its smart capabilities and the possibilities offered by Hyperledger Fabric channels.
Description
aWISH offers a cost-efficient solution to evaluate and improve the welfare of meat-producing livestock at a large scale, across Europe. At the heart of the aWISH solution is the automated monitoring at the slaughterhouse of complementary animal-based indicators for monitoring welfare onfarm, during (un)loading, transport and slaughter. Besides that, existing or routinely collected data (slaughterhouse data, antibiotics usage, farm data, etc.) and needed technologies on-farm or on-transport to complement the measurements at slaughter will be exploited. Novel sensor technologies and AI algorithms will be developed, and a feedback tool and interface will allow each actor in the chain to get direct feedback of each batch, visualize trends and benchmark animal welfare outcomes. An Animal Welfare Indicator Catalogue will disseminate all validated indicators and standardized data collection methods.Key Contributions
Within aWISH, UBITECH will develop data-driven tools for animal welfare (AW) assessment and improvement, through the development of a feedback and benchmarking tool for business operators (monitoring), Best Practive Guides - BPGs (improving), and assessing impact of certain practices on AW.
Description
COMFORTAGE is a joint effort of medical experts (i.e., neurologists, psychiatrists, neuropsychologists, nurses, memory clinics), social scientists and humanists, technical experts (i.e., data scientists, AI experts, robotic experts) and Digital Innovation Hubs to establish a pan European framework for Community-based, Integrated and People-Centric prevention, monitoring and progression managing solutions for age-related diseases and disabilities. The project’s framework will be empowered by a unique combination and integration of: (i) Medical/clinical innovations (e.g., novel approaches to risk factor analysis and personalized prediction, AI-based medical devices, integrated data sources of age-related clinical evidence, and evidence-based Healthcare Technology Assessment (HTA)), (ii) Cutting edge AI innovations (e.g., explainable AI (XAI), secure AI, serious games, Patient Digital Twins, Virtual Assistive technologies) for trusted, accurate, secure and personalized clinical decision making, (iii) Digital Innovation Hubs (e.g. Smart Homes, robotics and Living Labs) to facilitate and promote research activities in the health and wellbeing domain, and (iv) social innovations for promoting innovative views and co-creating new or improved solutions for assistance and improvement of social integration and interaction. COMFORTAGE will facilitate the integration, harmonization, and management of a host of different data sources, including biobanks, cohorts, medical records, longitudinal observational studies, real-world data about patients, as well of alternative secondary data sources, such as sensors, wearables and mobiles in a standardized structure called Holistic Health Records (HHRs). COMFORTAGE will become a catalyst to help prevent, monitor, and manage progression of age-related diseases and disabilities, especially of dementia and frailty, based on high-end research and analysis of the utilization of the aforementioned technologies.Key Contributions
Within COMFORTAGE, UBITECH will drive the delivery of the mechanisms that realise the concept of personalised and integrated Holistic Health Records (HHR), as well the key infrastructures (i.e., Blockchain, IKBs) and AI tools for the further utilization of the processed data. In particular, UBITECH will specify and implement the infrastructure required for the project’s ageing data space (i.e. Ageing-EHDS Infrastructure and Connectors). Specifically, this work will focus on the specification and implementation of an identify management infrastructure (e.g., based on KeyCloak), data models and ontologies for dementia and frailty related data, as well as connectors for contributing and consuming data to/from the data space. The work will be driven by the project’s DHRM and the implementation will be compliant to Reference Model of the IDSA and fully aligned to EHDS’s specifications. Interfaces and touch points for the integration to the wider EHDS will be specified as well. In addition, UBITECH will select, develop, and validate a range of AI algorithms for dementia and frailty prediction, prevention and delay of progression. The algorithms will include traditional ML techniques, Deep Learning algorithms (e.g., for medical imaging) and reinforcement learning (e.g., for identifying patients’ phenotypes). UBITECH will also implement an Integrated Care Model Library where the implemented models are stored and that offers relevant APIs to support the continuous enrichment/adaptation of development models based on the availability and analysis of more (risk, predictions, feedback) data.
Description
FAME is a joint effort of world-class experts in data management, data technologies, the data economy, and digital finance to develop, deploy and launch to the global market a unique, trustworthy, energy efficient, and secure federated data marketplace for Embedded Finance (EmFi). The FAME marketplace will alleviate the proclaimed limitations of centralized cloud marketplaces towards demonstrating the full potential of the data economy. In this direction, the project will enhance a state of the art data marketplace infrastructure (i.e., H2020 i3-Market marketplace) with novel functionalities in three complementary directions namely: Secure, interoperable, and regulatory compliant data exchange across multiple federated cloud-based data providers in-line with emerging European initiatives like GAIA-X; Decentralized, programmable, data assets trading and pricing leveraging blockchain tokenization techniques (including support for accruing data assets value in NFTs); Integration of trusted and Energy Efficient (EE) analytics based on novel technologies such as Quantitative Explainable AI, Situation Aware Explainability (SAX), incremental EE analytics, and edge analyticsKey Contributions
Within FAME, UBITECH implements a unified approach to managing and enforcing data policies in the FAME marketplace, supporting access to the security policies of the underlying data marketplaces and data spaces, along with mechanisms for their consolidation at the level of the FAME federated platform, and developing a security policies management tool, which will be able to map FAME policies to the lower level policies of the underlying providers.
Description
SEARCH is designed to overcome the traditional barriers in healthcare data sharing, such as privacy concerns, security risks, and institutional silos. By integrating synthetic data generation technologies with federated learning models, SEARCH ensures that sensitive patient data remains protected while fostering collaboration across public and private sectors. SEARCH will empower AI-driven clinical decision support and enable the development of novel healthcare tools, accelerating research, reducing bias, and improving the accuracy of diagnosis and treatment. SEARCH focuses on three critical areas of healthcare: cardiovascular, gastrointestinal, and gynecological diseases. We generate high-quality synthetic datasets that replicate real-world healthcare data while ensuring GDPR compliance and privacy protection. Our synthetic data mimics electronic health records (EHRs), genomics, imaging, and medical signal data, enabling healthcare professionals and researchers to conduct advanced analytics without compromising patient privacy. Through federated learning, SEARCH allows institutions to collaborate on AI/ML models without the need to share actual datasets. This not only protects sensitive information but also facilitates large-scale research that accelerates innovation in healthcare diagnostics, personalised treatment, and predictive modellingKey Contributions
Within SEARCH, UBITECH leads the design and implementation of the Data Harmonization Module, consolidating criteria and developing tools for data harmonization so as to contribute to increasing the comparability among data from different providers, vendors and data acquisition protocols, thus supporting data reusability in the context of statistical and AI-driven methods and promoting the development of reproducible and generalizable AI models. A range of approaches will be explored including prospectively harmonized acquisition, self- supervised learning and GAN networks. In addition to this, UBITECH contributes to the design of the Technical Platform Architecture, integrating under a common portal (with a single-entry point and a distributed architecture) the components and services to be developed. The main axes of the technical design will be a federated data network, providing data discoverability (using protocols such as GA4GH’s Beacon) and controlled access (making use of OpenID Connect), and a federated learning module, providing the necessary tools and procedures for building AI/ML solutions on such data networks. Moreover, UBITECH actively contributes to the specification of the technical set-up of the SEARCH federated nodes, supporting the technical establishment and federation of the local data warehouses at individual clinical sites and/or aggregated data repositories, as well as to the development of the Data Catalogue Federated Search and Discoverability, supporting data interoperability through an Extract-Transform-Load (ETL) toolset and extending synthetic dataset specifications to a metamodel that will describe synthetic datasets across cohorts from multiple angles, as well as to the design and development of the Federated Machine Learning Module, providing the data infrastructure so that the SEARCH platform modules can run distributed learning, with focus on dataset discoverability and access to controlled data.
Description
WATSON provides a methodological framework combined with a set of tools and systems that can detect and prevent fraudulent activities throughout the whole food chain thus accelerating the deployment of transparency solutions in the EU food systems. The proposed framework will improve sustainability of food chains by increasing food safety and reducing food fraud through systemic innovations that a) increase transparency in food supply chains through improved track-and-trace mechanisms containing accurate, time-relevant and untampered information for the food product throughout its whole journey, b) equip authorities and policy makers with data, knowledge and insights in order to have the complete situational awareness of the food chain and c) raise the consumer awareness on food safety and value, leading to the adoption of healthier lifestyles and the development of sustainable food ecosystems. WATSON implements an intelligence-based risk calculation approach to address the phenomenon of food fraud in a holistic way. The project includes three distinct pillars, namely, a) the identification of data gaps in the food chain, b) the provision of methods, processes and tools to detect and counter food fraud and c) the effective cross border collaboration of public authorities through accurate and trustworthy information sharing. WATSON will rely upon emerging technologies (AI, IoT, DLT, etc.) enabling transparency within supply chains through the development of a rigorous, traceability regime, and novel tools for rapid, non-invasive, on-the-spot analysis of food productsKey Contributions
Within WATSON, UBITECH implements an AI-supported Early Warning System for Detection and Prevention of Fraudulent Practices, for Food Safety Authorities based on the processing of various data and the connection with food fraud related databases. Moreover, UBITECH contributes towards the development of the blockchain-based data storage of WATSON platform and the related data usage/access smart contracts.Description
HIVE will carry out Implementation Research (IR) to support integrated care and management of HIV and concomitant NCDs leveraging an mHealth application tailored for the (self-)management of multiple long-term conditions for PLWH. This initiative will also incorporate interpersonal counselling (IPC) for vital psychosocial support and integration of NCDs care into existing HIV clinics. The mHealth application will build on an existing digital health tool developed by HIVE partner, Columbia University, which has already been piloted in several low- and middle-income countries (LMICs)7. HIVE aims to extend and customise the application for diverse high-income countries (HICs) and low- and middle-income countries (LMICs), including Kenya, Kazakhstan, Greece and Malta, addressing the unique needs of PLWH in different disease settings and cultural contexts. The goal is to mitigate the disparities in access to health for PLWH in different countries as well as the inequalities that are being faced among different population groups within each participating country. The HIVE mobile application will provide personalised health management strategies, including medication and appointment reminders, symptom tracking, educational material and lifestyle recommendations, all aimed at empowering PLWH to take an active role in managing their health. Additionally, the application will serve as telemedicine platform for IPC, facilitating access to mental and emotional health support to reduce social inequalities and stigmatisation. HIVE will collaborate with Global Alliance for Chronic Diseases (GACD), aligning its objectives with the Alliance's commitment to advancing implementation science for managing MLTCs and addressing NCDs in diverse and underserved populations.Key Contributions
Within HIVE, UBITECH will lead the Quality Control and Risk Management Task, ensuring that the project’s scientific and technical results are produced following high-quality standards and European guidelines, ensuring all activities during the project’s duration meet these benchmarks. A comprehensive Quality Assurance Plan (QAP), integrated into the Management Handbook, will be implemented to serve as a central reference for quality standards. In addition, UBITECH will contribute to the extension of HIVE Application to Manage MLTCs, redesigning the user interface to render it more user-friendly and intuitive, enriching the content to match the needs of the users, add features not available in its current version as described in the methodology, ensuring compatibility with the latest versions of operating systems, including more languages and integrating AI technology to deliver smart wellness recommendations and increased usability. Moreover, UBITECH will contribute to the cultural adaptation of the IPC guidelines and the contextualisation on HIV, including a comprehensive analysis of the available cultural adaptation frameworks for psychosocial interventions (e.g., EVF, CTAF, PAMF) to understand important elements of cultural adaptation and guide framework selection, and will also contribute to the conduction of assessments of cultural norms, societal attitudes toward mental health, and the mental health needs of the target populations to lead the adaptation. Last but not least, UBITECH will contribute to safeguarding HIVE's Security and Privacy Compliance, ensuring that the HIVE application meets regulatory standards such as GDPR to protect patient data, including the implementation of end-to-end encryption, role-based access controls, and secure authentication (e.g., MFA) to protect the data during storage and transmission.
Description
The BigDataOcean H2020-732310 project enables maritime big data scenarios for EU-based companies, organisations and scientists, through a multi-segment platform that will combine data of different velocity, variety and volume under an inter-linked, trusted, multilingual engine to produce a big-data repository of value and veracity back to the participants and local communities, bringing together the data, the network and the technologies to create a curated, semantically enhanced, interlinked and multilingual repository for maritime big data, where different stakeholders will be able to contribute with data in order to support their own goals and operations, but also allow new stakeholders (e.g. entrepreneurs, local communities, local authorities) will be able to develop new solutions in order to enable new, socially and environmentally sustainable business models or solutions.Key Contributions
Within BigDataOcean, UBITECH undertakes the technical integration lead of the project’s R&D activities, while UBITECH R&D team heavily contributes and leads the technological choices towards the definition and design of the integrated maritime data value chain, wherein all individual maritime-related stakeholders will be identified and brought together, bringing forward the value that one can add to each other, under a seamless collaboration and mutually beneficiary prism, adding stakeholders that priory seemed disconnected/irrelevant in order to realize a holistic big data-centred value chain. Moreover, UBITECH delivers the scalable architecture of the BigDataOcean framework enabling the integrated maritime data value chain, as well as the BigDataOcean libraries and algorithms for knowledge extraction, business intelligence and usage analytics, including (indicatively) clustering, segmentation, classification (e.g. linear regression, association rules and decision trees) in order to receive statistical data; constituting an extensible big (linked) data analytics apps library that will enable various types of maritime stakeholders to utilise and share analytic methods on big data for the discovery and communication of meaningful knowledge and new patterns that were unattainable or hidden in the previous isolated data structures; and targeting maritime-related stakeholders that usually do not tend blending their data, in order to setup business intelligence systems through the use of cross-sectorial and multi-lingual big data and gain significant competitive advantage.
Description
The ICARUS H2020-780792 project aims to build a novel data value chain in the aviation-related sectors towards data-driven innovation and collaboration across currently diversified and fragmented industry players, acting as multiplier of the “combined” data value that can be accrued, shared and traded, and rejuvenating the existing, increasingly non-linear models / processes in aviation. Using methods such as big data analytics, deep learning, semantic data enrichment, and blockchain powered data sharing, ICARUS will address critical barriers for the adoption of Big Data in the aviation industry (e.g. data fragmentation, data provenance, data licensing and ownership, data veracity), and will enable aviation-related big data scenarios for EU-based companies, organizations and scientists, through a multi-sided platform that will allow exploration, curation, integration and deep analysis of original, synthesized and derivative data characterized by different velocity, variety and volume in a trusted and fair manner. ICARUS will bring together the Aerospace, Tourism, Health, Security, Transport, Retail, Weather, and Public sectors and accelerate their data-driven collaboration under the prism of a novel aviation-driven data value chain. Representative use cases of the overall domain’s value chain include: (i) sophisticated passenger handling mechanisms and personalised services on ground facilities, (ii) enhanced routes analysis of aircrafts for improved fuel consumption optimisation and pollution awareness, (iii) more accurate and realistic prediction model of epidemics, and (iv) novel passenger experiences prein- and post- flight.Key Contributions
Within ICARUS, UBITECH will be responsible for the overall project management and consortium coordination, as well as for the architectural design and continuous integration of the final platform. Moreover, UBITECH R&D team will drive the design of a novel framework that leverages data, primary or secondarily related to the aviation domain, coming from diverse sources (data APIs, historical data, statistics, sensor / IoT data, weather data, and various other open data sources) to help companies and organisations whose operations are directly or indirectly linked to aviation to simultaneously enhance their data reach, as well as share and/or trade their existing data sources and intelligence, in order to gain better insights into airplanes’, airports’ and passengers’ quantified selves and contribute to improving their operations whether in real time or “offline” and increasing passengers’ safety and satisfaction. Finally, UBITECH will drive the implementation of (a) the Data Collection Services Bundle, mining, semantically annotating, transforming and checking in the collected data in the ICARUS data lake, ensuring the integrity and veracity of the data; (b) the Data Curation and Linking Services Bundle, encompassing data filtering, cleaning, normalization, modelling and mapping to a standardised data schema, translation, indexing and smart linking to existing datasets under linked open data repositories following the linked data principles; and (c) the Data Analytics Services Bundle that essentially correlates and analyses data to generate new insights and knowledge, incorporating typical machine learning techniques, such as regressions, support vector machines and k-means clustering, deep learning algorithms and prescriptive analytics.
Description
The INFINITECH H2020-856632 project is a joint effort of global leaders in ICT and finance towards lowering the barriers for Big Data, IoT and AI driven innovation, boosting regulatory compliance and stimulating additional investments and aims to provide novel Big Data and IoT technologies for seamless management and querying of all types of data (e.g., OLAP/OLTP, structured/unstructured/semi-structured, data streaming and data at rest), interoperable data analytics, blockchain-based data sharing, real-time analytics, as well as libraries of advanced AI algorithms; regulatory tools incorporating various data governance capabilities (e.g., anonymization, eIDAS integration) and facilitating compliance to regulations (e.g., PSD2, 4AMLD, MIFiD II); and nine novel and configurable testbeds and sandboxes, each one offering Open APIs and other resources for validating autonomous and personalized solutions, including a unique collection of data assets for finance and insurance. The results of the project will be validated through 14 pilot projects, which will cover all areas of reference for the sector, such as know your customer (KYC), customer analytics, personalised portfolio management, credit risk assessment, preventive financial crime analysis, fraud anticipation, usage-based insurance, agro-insurance and much more. INFINITECH will also make a platform available that will provide access to project solutions, along with a Virtualized Digital Innovation Hub (VDIH). This will support innovators such as FinTech and InsuranceTech in their ongoing commitment to the concrete application of Big Data technologies, AI and IoT.Key Contributions
Within INFINITECH, UBITECH will lead the development of a range of enablers for efficient, high-performance analytics that combine data from multiple sources and enable low-latency, near real-time operations. To facilitate high-performance analytics, UBITECH will contribute in the parallelization of incremental algorithms that are commonly used in finance and insurance applications (e.g., clustering, collaborative filtering and frequent pattern matching) as a means of accelerating their execution. Moreover, UBITECH will work towards the implementation of a declarative analytics framework that will enable analytics over diverse data sources based on conventional SQL-like primitives and in a way that handles the underlying complexity of data sources, as well as of a library of ML/DL algorithms for analytics in the finance and insurance sector, including both conventional algorithms and parallelized incremental algorithms. Finally, UBITECH will create a Big Data Workbench for use by the INFINITECH data scientists, for accessing the added-value analytics functionalities of INFINITECH such as the declarative analytics functionalities, as well as the ML/DL algorithms, and for executing analytics over the INFINITECH data management infrastructure and data analytics enablers.
Description
The SYNERGY H2020-872734 project introduces a novel reference big data architecture and platform that leverages data, primary or secondarily related to the electricity domain, coming from diverse sources (APIs, historical data, statistics, sensors/ IoT, weather, energy markets and various other open data sources) to help electricity stakeholders to simultaneously enhance their data reach, improve their internal intelligence on electricity-related optimization functions, while getting involved in novel data (intelligence) sharing and trading models, in order to shift individual decision-making at a collective intelligence level. To this end SYNERGY will develop a highly effective a Big Energy Data Platform and AI Analytics Marketplace, accompanied by big data-enabled applications for the totality of electricity value chain stakeholders (altogether integrated in the SYNERGY Big Data-driven EaaS Framework). SYNERGY will be validated in 5 large-scale demonstrators, in Greece, Spain, Austria, Finland and Croatia involving diverse actors and data sources, heterogeneous energy assets, varied voltage levels and network conditions and spanning different climatic, demographic and cultural characteristics.Key Contributions
Within SYNERGY, UBITECH drives the implementation of the End-to-end Interoperable Big Data Management Platform, incorporating the SYNERGY backbone infrastructure of the Core Big Data Management Platform (CBDMP), as well as the trusted data containers of the On-Premise Environment (OPE) and the Secure Experimentation Playground (SEP). Moreover, UBITECH will be responsible for the delivery of the data-at-rest and data-in-motion ingestion, management and curation services, as well as the end-to-end security, encryption and privacy assurance services in accordance with the requirements elicited for the energy domain.
Description
The XMANAI H2020-957362 project aims at placing the indisputable power of Explainable AI at the service of manufacturing and human progress, carving out a “human-centric”, trustful approach that is respectful of European values and principles, and adopting the mentality that “our AI is only as good as we are”. XMANAI will help the manufacturing value chain to shift towards the amplifying AI era by coupling (hybrid and graph) AI "glass box" models that are explainable to a "human-in-the-loop" and produce value-based explanations, with complex AI assets (data and models) management-sharing-security technologies to multiply the latent data value in a trusted manner, and targeted manufacturing apps to solve concrete manufacturing problems with high impact.Key Contributions
Within XMANAI, UBITECH implements the Security, Privacy and Trust Components for Industrial Asset Management, involving the Access Control Policy Manager, the Secure Data/Features and Models storage of the CAIMP, and the Secure Storage in the Secure Execution Clusters and the On-premise environments. Moreover, UBITECH leads the development of the Federated/On-Premise AI Models Execution and Visualization Environment, constituting a a containerised solution, replicating much of the components of the Secure Execution Cluster infrastructures, as shown in the architecture figure, which would be able to run on-premise and not being bound to any specific OS, allowing end users to have at their hands a solution that meet their needs, but that also communicates with the cloud based platform for acquiring data and utilising features that are not impacting in any way security, trust and computation performance (for example the data sharing methods etc). The environment to be delivered will allow the execution of AI model locally and the visualisation and extraction of results, while it will be able to export specific AI model configurations to be ported to systems that manufacturers already use, through the export of JSON files and of specific AI/Machine learning libraries that would be forked from existing repos of popular AI and machine learning and analytics frameworks and libraries, such as Spark, TensorFlow, DGL, Theano, Keras, Eurler, etc.IAM & Policy Manager
Description: An identity and access management (IAM) and policy enforcement layer designed to provide secure and flexible access control across data platforms. Built using Keycloak, it supports authentication, role-based access control (RBAC), attribute-based access control (ABAC), fine-grained data-level authorization, and single sign-on (SSO) integration. This solution enables centralized management of user identities and permissions to ensure compliance and robust security in diverse application environments.
Technologies & Features: Leverages Keycloak for centralized identity management, token-based authorization, and support for OAuth2/OpenID Connect protocols, with customizable access control policies. Built on Spring Boot (Java) for scalable microservice development and integration. Utilizes the EvalEx library for dynamic evaluation of policy expressions, enabling flexible and programmable access decisions.
Real-world Usage: Deployed to manage and control data access for analysts, data scientists, and applications, ensuring strict adherence to internal policies and regulatory requirements. These components are actively utilized in European research and innovation initiatives such as AWISH, XMANAI, and FAME, providing secure, standards-based identity and access management across distributed, data-intensive environments.
Use Cases & Domains: Useful in any domain requiring strict data governance—healthcare, public sector, finance, and education.
Unified Data Lakehouse Platform
Description: A scalable, high-performance data lakehouse architecture integrating Trino, Apache Iceberg, and ORC. The platform enables federated querying across heterogeneous sources, transactional consistency, schema evolution, and historical data access. Built to support both real-time and batch analytics, it reduces the complexity of managing disparate data systems.
Technologies & Features: Built using Trino (federated SQL query engine), Apache Iceberg (ACID-compliant table format), and ORC (optimized columnar storage), this platform offers powerful data management capabilities such as time travel, partition pruning, metadata indexing, high compression, and ANSI SQL compatibility across diverse data sources. It also integrates seamlessly with scalable object storage solutions like MinIO, enabling cost-effective, cloud-native storage and high availability for large-scale datasets.
Real-world Usage: The platform originated from the ICARUS project, providing a secure and policy-driven framework for managing distributed data sharing and collaboration. It has since evolved to support diverse applications, including its role in the aWISH project as a centralized system for collecting and analyzing data across the food production chain. This enables researchers and stakeholders to gain meaningful insights while ensuring data security, compliance, and traceability throughout the process.
Use Cases & Domains: Ideal for enterprise analytics, data warehousing modernization, governance-heavy industries (e.g., healthcare, financial services).
Google Fit Connector
Description: The Google Fit Connector module is a comprehensive data acquisition framework designed to seamlessly gather health and fitness information from a wide variety of wearable devices and fitness applications. Built on top of the Google Fit ecosystem, this artifact eliminates dependency on single-vendor ecosystems such as Fitbit or Garmin by leveraging Google Fit’s broad compatibility across multiple platforms. It facilitates the collection of diverse health metrics—such as activity levels, heart rate, and sleep patterns—aggregated through wearable devices and fitness apps synced with Google Fit. This architecture supports scalable, non-intrusive data collection, making it ideal for real-world health monitoring and AI-driven wellness applications.
Technologies & Features: Developed using Java and Spring Boot with OAuth2-based authentication, the module enables periodic data synchronization and historical backfill of health data
Real-world Usage: Used in wellness and health-monitoring apps to track user behavior, fitness trends, and population health. Developed as part of the ASCAPE EU project.
Use Cases & Domains: Relevant for digital health platforms, employee wellness programs, clinical research, and public health monitoring.
AI Security Engine
Description: The AI Security Engine provides a comprehensive security framework designed to protect AI models against a variety of adversarial threats. It focuses on three main areas: (a) identifying potential attacks such as pre-training poisoning, post-training evasion, and backdoor manipulations, (b) assessing the risks associated with these attacks, and (c) delivering detailed reports to users to enable informed corrective actions based on their expertise. This framework ensures the integrity and security of AI models throughout their lifecycle, enhancing their robustness and trustworthiness in deployment.
Technologies & Features: AI Security Engine is built using Python and leverages FastAPI to provide high-performance, scalable APIs for seamless integration with AI security functionalities. Kafka handles real-time event streaming, enabling prompt detection and response to security threats. PostgreSQL serves as the reliable backend database for managing detailed logs and assessment results. MinIO offers efficient, secure multi-cloud object storage for managing large datasets and model artifacts. Together, these mature technologies create a robust and extensible security framework for AI models.
Real-world Usage: The AI Security Engine is actively deployed in research and innovation projects such as XMANAI and is currently being enhanced within the AI-DAPT initiative. It supports AI practitioners by detecting and mitigating adversarial threats, thereby strengthening the security and reliability of AI models. Its use across diverse sectors contributes to increased trust and resilience in AI-driven systems.
Use Cases & Domains: The framework is highly relevant in sectors where AI model integrity is critical, such as autonomous systems, healthcare diagnostics, finance, cybersecurity, and critical infrastructure. Organizations deploying AI models in sensitive or high-stakes environments benefit from its ability to preemptively identify and mitigate security vulnerabilities, thereby ensuring safe and trustworthy AI operations.
Big Data & Predictive Analytics Framework
Description: The Big Data & Predictive Analytics Framework developed by the DEA Research Group enables robust and scalable handling of extensive data volumes. It incorporates high-performance predictive analytics methodologies, leveraging sophisticated machine learning (ML) and deep learning (DL) techniques to deliver accurate, timely insights for strategic decision-making. Specifically designed to support diverse industry needs, the framework allows seamless integration and analysis of structured and unstructured datasets.
Technologies & Features: The framework utilises cutting-edge technologies including PyTorch and TensorFlow for real-time data ingestion, model training, and inference. It supports advanced predictive models including transformer-based architectures and reinforcement learning agents, enabling the delivery of highly accurate predictive analytics solutions across time-series, tabular, and multimodal data. Integration with stream processing engines like Apache Flink, further enhances its real-time analytics capabilities.
Real-world Applications: This framework is actively deployed in ongoing Horizon Europe projects like COMFORTAGE, enhancing capabilities in predictive analytics across sectors such as healthcare, finance, agriculture, and food security.
Potential Use Cases & Beneficiaries: Ideal for predictive maintenance, fraud detection, market trend forecasting, and health outcome prediction, benefiting sectors including e-health, fintech, agriculture, and supply chain management.
Meta-learning Agentic AI
Description: The Meta-learning Agentic AI artefact focuses on adaptive artificial intelligence systems capable of rapid learning and generalisation across diverse tasks with minimal supervision. This artefact advances the state-of-the-art by integrating optimisation-based meta-learning techniques that facilitate quick adaptability in dynamic environments, crucial for real-world applications with limited labelled data.
Technologies & Features: Incorporates optimisation-based meta-learning methods such as Model-Agnostic Meta-Learning (MAML), metric-based techniques, and Bayesian meta-learning for uncertainty quantification. Implemented using modern ML libraries like PyTorch, learn2learn, higher, and Pyro, it supports automated meta-learning pipelines tailored to specific task requirements and cross-domain generalisation.
Real-world Applications: Employed in Horizon Europe projects including COMFORTAGE, enabling quick adaptation to new data for personalised healthcare monitoring, federated learning models, and fraud detection in complex supply chains.
Potential Use Cases & Beneficiaries: Highly applicable in areas such as healthcare diagnostics, personalised financial advisory systems, autonomous systems adaptation, and dynamic risk management, benefiting healthcare providers, financial institutions, logistics companies, and security agencies.
Group Leader
Dr. Konstantinos Perakis (Head of Group)
Expertise: (Big) Data Engineering, Biomedical Engineering, (Big) Data Management, Project and Fundraising Management
Short Bio
Dr. Konstantinos Perakis was born in Athens, Greece in 1979. He received his diploma in Electrical & Computer Engineering from the National Technical University of Athens in October 2003. He received his M.Sc. in Techno-Economical Systems in 2005 and his Ph.D degree in Medical Informatics in 2009. Since 2004 he has prepared numerous European proposals and has evaluated 3rd party proposals prior to their submission to the Commission, and has been active in a number of European and National R&D programs, through which he has gained considerable experience in the field of e-Health and m-Health, Cloud Computing & Cloud interoperability, Big Data, Data Analytics and Mining and Security. He has served as a Scientific and Technical coordinator in national projects. He has also served as an evaluator for the European Commission for co-funded R&D proposals, as well as a reviewer for the European Commission for co-funded R&D projects. Dr. Perakis has excellent communication and mediation skills, and the ability to deal well with people in many different contexts, which he has gained through his participation in multi-national consortia within the context of European projects. He has published more than 20 scientific papers in international journals and conferences as well as two book chapters, all in the field of Information and Telecommunications Technologies, and has served as reviewer and chairman in international journals and conferences.
Key Team Members
Dr. Aristidis Tsitiridis (Research Projects Manager)
Expertise: Computer vision, meta-learning, machine learning, AI systems, real-time data analysis, deep learning, healthcare AI, logistics AI, security AI, academic research, industrial innovation, EU-funded AI projects, research leadership.
Short Bio
Dr. Aristeidis Tsitiridis is an experienced researcher and professional in the fields of computer vision, machine learning and artificial intelligence, with an extensive educational background that includes a Ph.D. in Computer Vision from Cranfield University, an M.Sc. in Electronic Mobile Communications from the University of South Wales, and a B.Eng. in Electrical and Electronic Engineering from Leeds University. He is currently working as a Research Projects Manager at UBITECH in Athens, Greece, where he actively organises and contributes to H2020 AI Research Projects. In some of his previous roles, he worked as a Principal Computer Vision Researcher at Darvis Inc., and as a Principal Research Engineer at Huawei Technologies, tackling together with high-performance teams challenging computer vision problems for clients across several sectors (healthcare, logistics, warehousing, security). His past research experience also includes roles as a Senior Research Scientist at Electronic IDentification in Madrid, for real-time automated document-face verification, and academic positions as a Postdoctoral Researcher in Biometrics at Rey Juan Carlos University (Spain), and Research associate in computer vision at Swansea University (UK). Dr. Tsitiridis has contributed to numerous scientific publications, is an active reviewer in several journals and actively maintains academic affiliations with two institutions.
Dimitris Bouras (Big Data R&D Engineer, Research Group Tech Lead)
Expertise: Core Java, Spring Framework, Spring Integration, Apache Camel, SQL, Apache Kafka, Kafka Streams, Apache Flink, PostgreSQL, Oracle, MongoDB, Redis, Apache Airflow, Spring Batch, Docker, Kubernetes, GitLab CI, Grafana.
Short Bio
Dimitris Bouras holds an MSc in Information Systems and an MEng in Electronic Engineering (Communications) from The University of Sheffield. He has worked for 2 years as an IT Consultant for Cognity SA and later (2007-2013) as a Software Engineer for Forthnet R&D Department participating in National and European Research projects occupied with the development of web-based applications using Java Technologies. From December 2013 until November 2023, he worked as a Software Engineer for Neurocom SA working with polyglot data store environments focusing on the development of appropriate modules for processing data streams with real-time performance requirements. Within the framework of CoherentPaaS project (EC Contract No 611068) he researched how to address the particular needs of diverse queries and datasets combining different and multi-disciplinary data stores (columnar, key-value, relational, document stores, etc.) with flexibility. Furthermore, he worked on Neurocom’s commercial projects such as NOVA MSDP Platform which connects prepaid subscribers with the operator’s services and enables outbound notifications. Dimitris is a member of the Big Data Data Science and Analytics unit of UBITECH with the role of Big Data R&D Engineer. His main activity is the full participation in European or national co-funded research projects, in which he is responsible for researching and applying state-of-the-art (Big) Data or ML/DL oriented solutions to address specific needs in various domains. His primary interests lie in Big Data systems, scalable storage solutions, end-to-end data pipelines, and techniques for data anonymization and visualization.
George Mandilaras (Big Data Engineer)
Expertise: Big Data Engineering, Backend development, Microservice architecture, ETL pipelines, Streaming data processing, Data modeling Data-intensive applications.
Short Bio
Georgios Mandilaras was born in Rio, Greece, in 1994. He earned his bachelor’s degree in Informatics and Telecommunications from the National and Kapodistrian University of Athens (NKUA) in February 2019, followed by a master’s in Data Science and Information Technologies with a specialization in Big Data and AI in September 2021. Beginning his career as a Software Developer and Research Associate, he co-authored multiple publications on geospatial data processing in leading journals and conferences. Georgios later transitioned into a Data Engineer role at Ubitech, where he designed and implemented data-intensive systems, combining batch and streaming workflows, ETL pipelines, and large-scale data modeling. His work spans backend development with a strong focus on microservices and modular architectures. Finally, he has gained experience about aligning system design with product needs, creating solutions that are technically robust.
Stelios Georgalas (Software Engineer)
Expertise: Vue, React, Next Js, Java, SpringBoot, Python, FastAPI, Docker, CI/CD
Short Bio
Stylianos Georgalas is a versatile and accomplished Software Engineer with a strong academic foundation in Computer Engineering from the University of the Peloponnese and a postgraduate specialization in Artificial Intelligence and Smart Software Applications from the University of Piraeus. With over six years of hands-on experience, he has successfully contributed to a wide range of R&D and commercial initiatives, including European funded projects and the development of sophisticated software platforms. His expertise spans software engineering, intelligent systems, and data driven application development, positioning him as a valuable asset in both an academic and industrial environment. Stylianos consistently demonstrates a capacity for innovation, leveraging his deep understanding of AI, system architecture, and scalable solutions to tackle complex engineering challenges with precision and creativity.
George Klados (AI Engineer)
Expertise: Machine Learning, Deep Learning, Artificial Intelligence, Early Warning Systems, Explainability, Biomedical signal processing, Biomedical image processing
Short Bio
Georgios Klados is a highly qualified researcher specializing in Big Data and Machine Learning, with particular emphasis on machine learning security and the development of scalable machine and deep learning pipelines. His interdisciplinary expertise encompasses software development, biomedical engineering, and data-driven research methodologies, positioning him as a valuable contributor to complex, high-impact scientific projects. He holds a Diploma in Electrical and Computer Engineering from the Technical University of Crete, where his undergraduate thesis focused on the registration of fMRI data onto a 3D brain atlas to support neurosurgical planning. His Master’s thesis is related to Biomedical Engineering and Machine Learning, during which he developed advanced methodologies for detecting high-frequency oscillations in intracranial EEG signals using tensor decomposition and machine learning techniques. Mr. Klados has actively participated in national and European research projects addressing critical biomedical challenges, including cardiovascular risk assessment and the detection of sleep apnea episodes. Additionally, his research contributions extend to the agri-food sector, particularly in the detection of fraudulent activities. He has presented his work at internationally recognized conferences and has demonstrated a strong capacity for interdisciplinary collaboration, working effectively with professionals across clinical, biological, and technical domains.
Nikos Chachampis (Senior Software / ML Engineer)
Expertise: Python, Automation, Testing, Cloud, Machine Learning, CI / CD, Big Data.
Short Bio
NIKOS CHACHAMPIS, Senior Software / ML Engineer, has a versatile engineering background spanning Cloud Computing, Data Engineering, and Machine Learning. He brings a unique blend of technical expertise and innovative problem-solving across telecom, mobility, and startup ecosystems. Holding the AWS Certified Solutions Architect - Associate certification, he leverages extensive experience in AWS, Kubernetes and advanced technologies like Large Language Models and ETL pipelines to consistently deliver complex technical solutions — from developing machine learning models for investment opportunities to orchestrating data pipelines processing terabytes of information. His academic foundation in Computer Engineering and Digital Signal Processing as well as his commitment to continuous learning underscore his technological advancement and versatility in the ever-evolving tech landscape.
Recent Highlights
Publications
-
Perakis K., Rodriguez Del Rey S., Lampathaki F., An Explainable & Secure Artificial Intelligence platform for the manufacturing industry – Applications & lessons learnt, 12th Int. Conf. on Interoperability for Enterprise Systems and Applications, I-ESA 2024, April 10th-12th 2024, Crete, Greece.
-
Koussouris S., Dalamagas T., Figueiras P., Pallis G., Bountouni N., Gkolemis V., Perakis K. et. al., "Bridging Data and AIOps for Future AI Advancements with Human-in-the-Loop. The AI-DAPT Concept," 2024 IEEE International Conference on Engineering, Technology, and Innovation (ICE/ITMC), Funchal, Portugal, 2024, pp. 1-8, doi: 10.1109/ICE/ITMC61926.2024.10794334
-
Agostinho C, Dikopoulou Z, Lavasa E, Perakis K, Pitsios S, Branco R, Reji S, Hetterich J, Biliri E, Lampathaki F, Rodríguez Del Rey S and Gkolemis V, 2023, Explainability as the key ingredient for AI adoption in Industry 5.0 settings. Front. Artif. Intell. 6:1264372. doi: 10.3389/frai.2023.1264372.
-
Pervanidou P., Chatzidaki E., Nicolaides N., Voutetakis A., Polychronaki N., Chioti V., Kitani R. A., Kyrkopoulou E., Zarkogianni K., Kalafatis E., Mitsis K., Perakis K., Nikita K., and Kanaka-Gantenbein C., The impact of the ENDORSE digital weight management pro-gram on the metabolic profile of overweight and obese children and on food parenting practices, Nutrients 2023, 15(7), Nutrients 2023, 15(7), 1777, https://doi.org/10.3390/nu15071777 .
Zarkogianni K., Chatzidaki E., Polychronaki N., Kalafatis E., Nicolaides N., Voutetakis A., Chioti V., Kitani R.A., Mitsis K., Perakis K. et. al., The ENDORSE Feasibility Study: Exploring the Use of M-Health, Artificial Intelligence, and Serious Games for the Management of Childhood Obesity, Nutrients 2023, 15(6), 1451; https://doi.org/10.3390/nu15061451
-
Mavrogiorgou A., Kiourtis A., Makridis G., Perakis K. et. al., FAME: Federated Decentralized Trusted Data Marketplace for Embedded Finance, 2023 International Conference on Smart Applications, Communications and Networking (SmartNets), 25-27 July 2023, 10.1109/SmartNets58706.2023.10215814.
-
Miltiadou D., Perakis K., Sesana M., Calabresi M., Lampathaki F., Biliri E., A novel Explainable Artificial Intelligence and secure Artificial Intelligence asset sharing platform for the manufacturing industry, 29th ICE IEEE/ITMC Conference (ICE 2023), 19-22 June 2023, Edimburgh, Scotland, https://doi.org/10.5281/zenodo.8010282
-
Serrano M., Khorsand B., Soldatos J., Troiano E., Neises J., Kranas P., Perakis K., et. al., INFINITECH Book Series – Part 1. Concepts and Design Thinking Innovation Addressing the Global Financial Needs: The INFINITECH Way Foundations. Boston–Delft: Now Publishers, 2023, ISBN: 978-1-63828-228-0, DOI: 10.1561/9781638282297
-
Lampathaki F., Biliri E., Tsitsanis T., Tsatsakis K., Miltiadou D., and Perakis K., Toward an Energy Data Platform Design: Challenges and Perspectives from the SYNERGY Big Data Platform and AI Analytics Marketplace, Data Spaces - Design, Deployment and Future Directions, pp. 293 - 315, Springer, 2022, ISBN 978-3-030-98635-3, https://doi.org/10.1007/978-3-030-98636-0.
-
Miltiadou D., Pitsios S., Perakis K., et. al., Leveraging Management of Customers’ Consent Exploiting the Benefits of Blockchain Technology Towards Secure Data Sharing, Big Data and Artificial Intelligence in Digital Finance, 2022, https://doi.org/10.1007/978-3-030-94590-9_8
-
Lampathaki F., Biliri E., Tsitsanis T., Tsatsakis K., Miltiadou D. and Perakis K., Toward an Energy Data Platform Design: Challenges and Perspectives from the SYNERGY Big Data Platform and AI Analytics Marketplace, Data Spaces - Design, Deployment and Future Directions, 2022, ISBN: 978-3-030-98638-4, https://doi.org/10.1007/978-3-030-98636-0
-
Miltiadou D., Pitsios S., Spyropoulos D., Alexandrou D., Lampathaki F., Messina D. and Perakis K., A Secure Experimentation Sandbox for the Design and Execution of Trusted and Secure Analytics in the Aviation Domain, International Conference on Security and Privacy in New Computing Environments, SPNCE 2020: Security and Privacy in New Computing Environments, pp. 120-134, 22 January 2021, doi: 10.1007/978-3-030-66922-5_8.
-
Miltiadou D., Pitsios S., Spyropoulos D., Alexandrou D., Lampathaki F., Messina D., Perakis K., A Big Data Intelligence Marketplace and Secure Analytics Experimentation Platform for the Aviation Industry, 10th EAI International Conference, BDTA 2020, December 2020, https://doi.org/10.1007/978-3-030-72802-1_4
- Perakis K., Lampathaki F., et. al., CYBELE – Fostering Precision Agriculture & Livestock Farming Through Secure Access to Large-Scale HPC Enabled Virtual Industrial Experimentation Environments Fostering Scalable Big Data Analytics, Computer Networks, Special Issue on The Big Data Era in IoT-enabled Smart Farming: Re-defining Systems, Tools, and Techniques, Elsevier, Volume 168, February 2020, https://doi.org/10.1016/j.comnet.2019.107035
For collaboration opportunities and other inquiries, contact us at [email protected]
