Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report

Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint insights from the Horizon Europe security research projects: FLEXI-cross, TENSOR, ODYSSEUS, and TENACITy Joint Report, August 2025 Co-funded by the European Union Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 2 Abstract This document presents key insights from four EU-funded projects—FLEXI-cross, TENSOR, ODYSSEUS, and TENACITy—working on border management, travel security, and law enforcement intelligence support. It highlights common challenges, lessons learned, and gaps identified across technical and operational implementations. This joint report serves as a roadmap for policymakers, security practitioners, industry stakeholders, and technology experts, fostering a more secure and intelligent approach to border management and law enforcement. Authors FLEXI-cross Authors Giuseppe Vella – Engineering Ingegneria Informatica S.p.A. (ENG) Marco Paleari – Engineering Ingegneria Informatica S.p.A. (ENG) Maria Giuseppa Guella – Engineering Ingegneria Informatica S.p.A. (ENG) Chrostos Bolakis – Kentro Meleton Asfaleias (KEMEA) Thomas Azrak – Ebos Technologies Limited– (EBOS) TENSOR Authors Eleni Veroni – Netcompany S.A. Spyridon Evangelatos – Netcompany S.A. Apostolos Apostolaras – Centre for Research and Technology, Hellas (CERTH) Katerina Kyriakou – Centre for Research and Technology, Hellas (CERTH) Thanasis Korakis – Centre for Research and Technology, Hellas (CERTH) Patrik Gonçalves – Central Office for Information Technology in the Security Sector (ZITiS) Tabea Rosenkranz – Central Office for Information Technology in the Security Sector (ZITiS) Claudia Mertinger – Fsas Technologies GmbH ODYSSEUS Authors Monica Florea – Software Imagination & Vision Romania (SIM) Dana Oniga– Software Imagination & Vision Romania (SIM) Diana Antonescu – Software Imagination & Vision Romania (SIM) Dimitris Kassimis – TELESTO TECHNOLOGIES PLIROFORIKIS KAI EPIKOINONION EPE (TEL) Martin David – THALES DIS CZECH REPUBLIC SRO (THALES) Harry Kellett – RAPISCAN SYSTEMS LIMITED (RAPI) TEANCITy Authors Chrysostomos Antoniou – European Dynamics Luxembourg SA (ED) Christiana Aposkiti – Kentro Meleton Asfaleias (KEMEA) Mirela Rosgova – Kentro Meleton Asfaleias (KEMEA) Deborah Manzi – Universita Cattolica Del Sacro Cuore (UCSC-TC) Celia Calus – Nutcracker Research Malta Ltd (NMT) Rodoula Makri – Institute Of Communication & Computer Systems (ICCS) Credits Design by Netcompany S.A. Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 3 Table of Contents 1. Brief introduction of the four projects ...........................................................................................4 1.1. FLEXI-cross introduction ..........................................................................................................................4 1.2. TENSOR introduction ..................................................................................................................................5 1.3. ODYSSEUS introduction .............................................................................................................................6 1.4. TENACITy introduction ............................................................................................................................... 7 2. Challenges ...............................................................................................................................................................9 2.1. Challenge 1: Biometric on the Move for seamless border crossing ..................................9 2.2. Challenge 2: Unobtrusive technologies for border crossing facilitation of people and goods .....................................................................................................................................9 2.3. Challenge 3: Modern biometric technologies in forensic science  forenhancedsuspectidentification ............................................................................................... 10 2.4. Challenge 4: Travel intelligence applied to FCT ......................................................................... 10 3. Use cases on AI applied to Biometric Authentication, Travel intelligence, and goods and people crossing borders .................................... 12 3.1. FLEXI-cross use cases ............................................................................................................................. 12 3.1.1. Preliminary results ................................................................................................................................. 14 3.1.2. Lessons learnt ..........................................................................................................................................14 3.1.3. Gap Analysis ..............................................................................................................................................18 3.1.4. Recommendations on possible standardisation activities and strategies for policy experts ................................................................................................. 20 3.2. TENSOR use cases ...................................................................................................................................... 22 3.2.1. Lessons learnt ......................................................................................................................................... 23 3.2.2. Gap Analysis .............................................................................................................................................26 3.2.3. Recommendations on possible standardization activities and strategies for policy experts ...................................................................................................31 3.3. ODYSSEUS use cases ................................................................................................................................. 33 3.3.1. Lessons learnt ......................................................................................................................................... 33 3.3.2. Gap Analysis ............................................................................................................................................ 37 3.3.3. Recommendations on possible standardization activities and strategies for policy experts .................................................................................................. 39 3.4. TENACITy use cases ...................................................................................................................................40 3.4.1. Lessons learnt ......................................................................................................................................... 42 3.4.2. Gap Analysis ............................................................................................................................................ 44 3.4.3. Recommendations on possible standardization activities and strategies for policy experts .................................................................................................. 46 4. Conclusions .........................................................................................................................................................48 4.1. Commonalities.............................................................................................................................................48 4.2. Needs ................................................................................................................................................................ 48 5. Acknowledgements .....................................................................................................................................50 Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 4 1. Brief introduction of the four projects 1.1. FLEXI-cross introduction The FLEXI-cross project aims to increase security and reliability of EU border checks for people and goods through the development, deployment and validation of a toolkit of innovative border-checking solutions. The resulting flexibility and dynamicity of border check planning will offer novel capabilities such as the dynamic deployment of checkpoints and support via mobile applications for border personnel while guaranteeing a high level of security, privacy of personal data and protection of people’s fundamental rights. Figure 1: FLEXI-cross outcomes Foreseeable Outcomes: • anti-trafficking and anti-smuggling protection via predictive risk assessment of vehicle and people • enhanced border security through portable biometric-based checks • secure person verification through real-time multi-source cross-referencing • flexible, fast and cost-effective deployment of ad-hoc Border Check Points • secure, private and traceable sensitive / personal data exchange based on blockchain technology • increased safety and improved experience for border personnel based on advanced Human Machine Interfaces and enhanced situational awareness via Augmented Reality Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 5 1.2. TENSOR introduction The TENSOR project aims to revolutionize the way Law Enforcement Agencies (LEAs) across Europe identify suspects and combat serious crime and terrorism. In an era of increasingly sophisticated threats and fragmented investigative workflows, TENSOR provides innovative, AI-powered solutions for the extraction, analysis, storage, and secure cross-border sharing of biometric evidence. Figure 2: TENSOR outcomes Foreseeable Outcomes: • Fusion of Physiological and Behavioural Biometrics: Advanced multi-modal biometric identification system that supports law enforcement agencies in identifying suspects for combating crime and terrorism more effectively. The platform integrates a broad spectrum of biometric modalities, combining well-established methods, such as fingerprints and facial recognition, with emerging behavioural biometrics, including gait recognition (based on unique walking patterns) and keystroke dynamics (based on typing behaviour such as timing, rhythm, and pressure). • Intelligence from Soft Biometrics: Derives a rich set of soft biometric traits, such as age, gender, body proportions, and language use, from multiple modalities to enrich suspect profiling and support cross-checking when hard biometric matches are unavailable. • Explainability for Fair and Unbiased Biometric Technologies: Embedded explainability mechanisms and expert-in-the-loop workflows, enhancing the system’s decision-making pipeline. • Biometrics Dataspace: Enables suspect identification in transnational cases by performing secure biometric data matching (e.g., facial images, voiceprints, fingerprints) across jurisdictions without exposing sensitive information. Built on IDSA data space technology, TENSOR uses privacy-enhancing technologies like homomorphic encryption to preserve privacy during processing, and blockchain-based smart contracts for automated access control and lawful data use. • Digital Forensics: Develops methodologies for extracting data of interest (e.g., app usage patterns, location traces, configuration data from health apps, media files, text messages, etc.) from seized mobile devices, bypassing the inherent encryption mechanisms. Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 6 • AI-powered Digital Assistant: Features an AI-powered Digital Assistant to support investigators, which consolidates biometric identification results into structured, natural language reports. Through a chat-based interface, the assistant explains how TENSOR’s technologies work, helps users navigate the platform, and provides step-by-step guidance for key investigative tasks. 1.3. ODYSSEUS introduction ODYSSEUS project represents a forward-looking response to the complex challenges European border management is facing. Funded under the Horizon Europe programme, ODYSSEUS is developing an integrated, and ethical border control platform that leverages cutting-edge technologies such as artificial intelligence (AI), machine learning (ML), and advanced biometric systems. Its mission is to enhance the security, efficiency, and scalability of border operations while respecting fundamental rights and ensuring compliance with EU regulations. As migration pressures rise and security threats evolve, border authorities are confronted with the dual challenge of maintaining operational effectiveness and respecting privacy and individual rights. ODYSSEUS addresses this by designing solutions that are unobtrusive, user-friendly, and adaptable to diverse border contexts—land, sea, and train. Through the deployment of UAV-assisted vehicle scanning, AI-based risk assessment engines, multi-sensor data fusion, and digital travel credentials, ODYSSEUS platform provides a comprehensive framework for next-generation border control. ODYSSEUS AI-powered border control platform Digital Travel Credentials (DTCs) UAV-assisted vehicle scanning X-ray integration for non-intrusive inspections Digital Wallet Decision Support System (DSS) Validation across multiple environments Figure 3: ODYSSEUS outcomes Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 7 Foreseeable Outcomes: The ODYSSEUS project will deliver a series of innovative results that contribute to the future of secure, efficient, and ethical border management across the EU. Key outcomes include: • AI-powered border control platform: A modular and scalable solution that integrates biometric verification, risk assessment, and surveillance into a unified system. • Digital Travel Credentials (DTCs): Implementation of secure and privacy-respecting credentials that facilitate faster and more reliable traveller identification. • UAV-assisted vehicle scanning: Deployment of unmanned aerial vehicles for license plate recognition, occupancy detection, and anomaly identification, enhancing situational awareness without disrupting traffic flow. • X-ray integration for non-intrusive inspections: Use of AI to interpret high-resolution X-ray images for vehicle and train screening, detecting threats or contraband in real-time. • Digital Wallet: A secure platform for travellers to manage and share personal and travel-related documentation, supporting GDPR compliance and minimizing manual checks. • Decision Support System (DSS): A tool that synthesizes multi-source data to support real-time decision-making and risk profiling by border authorities. Validation of these innovative solutions for border crossing systems at European level will improve the border crossing experience for travellers and for border guards. The evaluation and testing of the ODYSSEUS platform are performed in land, sea, and rail scenarios, demonstrating flexibility and real-world applicability for EU internal and external borders. The security and reliability of border checks will be increased through new solutions of identification of people and goods crossing external borders, while protecting people’s fundamental rights and personal data. These outcomes will collectively support a more resilient, interoperable, and citizen-centric border management approach aligned with EU regulations and its strategic objectives. 1.4. TENACITy introduction Law Enforcement Agencies (LEAs) across Europe rely on data from their information systems to make critical decisions that ensure the safety of European citizens. However, a recent report from the European Court of Auditors has revealed inconsistencies in how individual countries manage and utilize data. Differences in perception and methodology have led to incomplete, inaccurate, and outdated datasets being used by LEAs, with some countries not fully leveraging the central EU systems due to regulatory and “cultural” barriers. The TENACITy project aims to address these challenges through a comprehensive three-pillar approach. Firstly, it proposes modern and effective tools for travel intelligence data utilization, building an interoperable open architecture that integrates and analyses multiple transactional, historical, and behavioural data from diverse sources. By leveraging cutting-edge digital technologies, such as machine learning, AI, blockchain, and distributed ledger technologies, TENACITy ensures secure and trusted information sharing between authorities. Secondly, it emphasizes the training and sensitization of LEAs’ personnel by establishing a living lab to organize hackathons, workshops, and interactive sessions for stakeholders, fostering knowledge exchange and practical application of the proposed digital tools. Finally, TENACITy envisions a holistic approach to crime prevention through the development of a Travel Intelligence Governance Framework that balances security with fundamental rights. By involving citizens, civil actors, and policymakers, the project addresses legal, ethical, and societal concerns while strengthening security measures against criminal and terrorist organizations. Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 8 Foreseeable Outcomes: • Advanced Travel Intelligence Utilization: Game-changing technologies to exploit global travel intelligence data, using machine learning and AI to extract patterns and eliminate irrelevant data. These technologies provide updates about potential threats, the trends of enterprises of crime and will thus strengthen their intelligence, analytic capacity and decision-making. • Trusted and Secure Information Sharing: Utilization of blockchain and distributed ledger technologies to securely share encrypted data among authorities, enabling transparent and traceable information exchange while protecting sensitive data. • Enhanced Decision-Making for Security Authorities: Real-time analytics and risk management algorithms consolidate data from various sources, boosting intelligence capacity and providing updates on emerging threats and criminal trends. • Dynamic Training and Stakeholder Engagement: Continuous training programs for law enforcement agencies (LEAs) through living labs, workshops, and hackathons to ensure thorough understanding and application of travel intelligence. Promotes collaboration and knowledge exchange among stakeholders. Holistic Crime Prevention Approach: Development of a Travel Intelligence Governance Framework to address legal, ethical, and societal concerns. Involves citizens, policymakers, and civil actors in shaping regulations to balance security with fundamental rights. Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 9 2. Challenges 2.1. Challenge 1: Biometric on the Move for seamless border crossing The capabilities to capture and use the biometrics of travellers without them having to stop and in natural contexts for border checks, in full respect of fundamental rights and considerations to safeguard data and integrity, are crucial aspects linked to the border crossing system innovation. New initiatives to accelerate the border crossing process are launched every year, introducing emerging technologies besides the established biometric methods, such as face recognition, since it is nonintrusive, fast, easy to use, and cost-effective. On the other hand, face recognition/identification, particularly in open environments and uncontrolled conditions, is a highly challenging task due to diverse variations of face appearance, face sizes, cluttered and complicated backgrounds, etc. and due to technical challenges, such as blurriness, low resolution, and acquisition conditions. An additional challenge is the efficient verification of travel documents in the case of minors, as their biometric features change rapidly until adulthood, thus making it impossible to verify the legal guardianship between them and the people accompanying them. To address the above challenges, the FLEXI-cross project implements diverse biometrics-solutions in the respective trial sites. A two-step approach has been followed for the construction of a deep feature representation database. The first step applies Face Detection and Face alignment techniques based on deep Convolutional Neural Networks (CNNs), while the second utilises face embedding learning and deep face feature extraction. Special emphasis has been given in increasing the accuracy and trustworthiness of the results, the usability and adaptation of the solution to real operational border environments. 5G connected portable devices will be used for multimodal biometrics to allow Law Enforcement Agencies (LEAs) to verify anyone in their territory with different databases in real time. Machine learning will be used to improve the system for the effective detection of people crossing borders. 2.2. Challenge 2: Unobtrusive technologies for border crossing facilitation of people and goods One of the main challenges concerning EU external borders in 2024 was the irregular migration according to the latest “Annual Risk Analysis 2024/2025” published by Frontex in July 20241. Even if the “Annual brief for 2024” issued by Frontex in February 2025 states that detections of illegal border-crossing on entry at the EU’s external borders decreased by 38% compared in 2024 compared with 20232, the irregular migration remains a risk for the EU border management. Another big challenge at EU external borders is detecting illegal activity without creating delays for other travellers, i.e. detecting false documents or smuggling illegal goods. ODYSSEUS solutions are meant to support border guards’ authorities to fight these challenges, while ensuring a smooth crossing for travellers across land borderson the road and on railway and on sea borders of EU. 1 Annual Risk Analysis 2024/2025, Frontex, July 2024. 2 Annual brief for 2024,Frontex,February 2025. Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 16 ID Name Responses 1anonymous We’ve created a common data model to be used by all partners for the generation of high level events allowing for more flexibility to the internal architectures. We’ve built a data validation tool and provided extensive feedback to all partners producing data. We’ve set several deadlines and provided additional help and guidance to partners. 2anonymous We defined a common data model for all partner to exchange events with the FLEXI-cross toolkit and a tool to allow technical partner to test event’s production. We implemented ETL procedures to allow ‘translation’ of messages to different data-formats. 3anonymous Made focus meeting with some, available BCP personnel to align the dashboard with what they could be looking for within their daily operation 4anonymous Keeping track of open points; Short term and long term planning of achievements and challenges early on; Integration meetings between partners to overcome challenges/ blockers. Conducted Workshops/trainings and collaborated on overall solution. 5anonymous Continuous testing and refinements 6anonymous Worked with legacy systems to build the necessary modules. Installed additional equipment at borders to achieve the envisioned functionality Figure 5: actions to overcome the challenges Q7. Lessons Learnt from Technical Challenges (actual responses reported in Figure 6) • Developing Compatibility Solutions: Innovative methods were required to bridge compatibility gaps between legacy systems and new technologies. • Decentralised System Adaptation: New methodologies were designed to accommodate non-standardised and decentralised systems. • Enhancing User Engagement: Special strategies were implemented to improve communication and training for non-technical end users. Figure 6: experience and lessons learned Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 17 Q15. Challenges Faced by End Users (BCPs) (actual responses reported in Figure 7) a) Bureaucratic and Administrative Barriers 66% of end users cited difficulties related to bureaucratic hurdles, such as delays in approvals, restricted data access, and coordination challenges between multiple stakeholders. b) Insufficient Training and Support 50% of participants mentioned that they lacked adequate training on new tools, leaving them unprepared for adoption. Many BCPs struggled to integrate new technologies into their workflow due to the absence of structured training sessions. c) High Workload Constraints The demanding nature of BCP operations left personnel with little time to dedicate to project-related activities or research and development efforts. Figure 7: Challenges reported by end users Q16. Experiences Gained and Positive Outcomes (actual responses reported in Figure 8) Despite these challenges, the experience gained throughout the project was overwhelmingly positive. All participating end-user partners reported: • Exposure to new technologies for the first time, expanding their awareness of digital solutions that could enhance border control operations. • A better understanding of how technology can support security operations without compromising efficiency. • Excitement about future collaboration between BCPs and the private sector, recognising the value of innovation in border security. Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 18 Figure 8: experience and lessons learned by end users 3.1.3. Gap Analysis A thorough gap analysis was conducted, centred on identifying inefficiencies and areas for improvement in operational workflows, ensuring alignment with strategic objectives and technological advancements. The work was conducted in two phases: • Phase 1: Definition of the framework for redesigning operational processes. • Phase 2: Definition of the anticipated complexity of executing these processes, based on the Key Performance Indicators explicitly identified and described in WP3, Trials and Validation. During Phase 1, a three-step methodology was adopted, as shown below: • Step 1: All of the end users’ functional legacy systems were considered while thoroughly defining the current procedures being followed. • Step 2: All suggestions for technological, operational, and other improvements were emphasised based on the identification of real-world gaps and direct extraction of end-users’ experience. • Step 3: The procedures were redesigned to enhance efficiency and effectiveness in terms of time, security, cost, and quality while checking passengers crossing the border points of interest. Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 19 An indicative result from Phase 1 related to UC3 is presented in the table below. Existing Operational Process (Only for third-country nationals) Technology used [Only Legacy systems] Operational Gap or/and proposed optimised Processes Redesign of the technological steps with time prioritisation, adding the tools of the FLEXI – cross solution Identity check and verification Visa or residence permit check Duration of authorised stay check Departure, destination and purpose verification Means of subsistence verification Means of transport and carrying objects check 1. CREDENCE ONEMRZ travel documents reader and verification, fingerprint scanner and verification 2. Visual and/or physical check by border guard 1. Biometric sensors, Document readers 2. 5G Communication infrastructure 3. Applications for authorities installed in relevant devices (smartphones) 4. Train wagons Thermal cameras Video cameras, acoustic sensors 5. Smartphones 5G 1. Installation of cameras and acoustic sensors 2. Implementing Biometric detection tool 3. Match with uploaded documents 4. Functionality checking. 5. Application of ML Algorithms on received data 6. Results Analysis and validation 7. Testing 5G network reliability 8. Integration in Control UI, where data monitoring takes place 9. Functionality validation on authorities’ devices During Phase 2, a methodological approach was applied to define complexity factors for KPIs in relation to process redesign. This allowed for the anticipation of an improved solution toolkit, with a focus on potential practical challenges during trial execution, particularly for KPIs with high complexity factors. This method was applied to the three use cases of interest, independently assessing the complexity of each KPI. Specifically, after determining that the Complexity Factor (CF) is defined by the number of redesigned steps included for each KPI, the trial and technical leaders of each use case assigned CF values to all KPIs across the three use cases. Finally, by averaging the complexity factors per use case, the results are presented in the figure below (Figure 9). Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 20 Figure 9: The averaged complexity factors (in percentages) for the three use cases of FLEXI-CROSS. 3.1.4. Recommendations on possible standardisation activities and strategies for policy experts This section shows how recommendations or evidence have been collected in order to provide EC officers with some good practices that have been experienced during the FLEXI-cross project. Several methods have been followed since the beginning of the project to provide the aforementioned evidence. First, a multi-topic set of activities, like synergy webinars and workshops, was organized to gather ideas from different contexts. Secondly, the participation to topic-specific events, like the European Association for Biometrics events, has been very useful to discuss, in a European context, about the main challenges and experiences of the different European actors. The main outcomes of these events have been minutes and the collection of common results within the same topic or between European Research topics like Fight Crime and Terrorism and Border Management. 3.1.4.1. Recommendations • Recommendation 1 – Projects Promotion • Recommendation 2 – Joint Exploitation and Dissemination • Recommendation 3 – Ethical and legal aspects curation • Recommendation 4 – Standardization groups and working group promotion Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 21 Figure 10: Recommendations overview. Recommendation 1 – Projects Promotion Aim 1: to intensively promote projects clustering activities and participation to synergy activities. Aim 2: To actively promote the main results of the projects in order to identify scientific and innovation findings. Where and Who was involved: EAB RPC conference, other Biometric conferences, Projects to Policy Seminar, CERIS, brokerage events. Target audience: RTO, Academies, Large industries, EC officers and agencies. Recommendation 2 – Joint Exploitation and Dissemination Aim 1: to promote European services that support joint exploitation and dissemination. An additional advertisement is needed by increasing awareness of Horizon Results Booster and the related tools. Projects are stimulated to think more in the context of shared results, and possible impacts. Where and Who was involved: PPS, Policy Officers and Project officers Target audience: RTO, Academies, Large industries, EC officers and agencies, projects facilitating the task. Recommendation 3 – Ethical and legal aspects curation Aim 1: to create more confrontation moments on the AI Act and its application to the Innovation and Research projects. Aim 2: to explore different nuances of research activities in order to cover with round tables and FAQ boards all the doubts related to the implementation of AI in the different contexts. Where and Who was involved: EAB RPC conference, Ethical Experts, EC legal and ethical experts. Target audience: RTO, Academies, Large industries, EC officers and agencies, projects facilitating the task. Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 22 Recommendation 4 – Standardization groups and working group promotion Aim 1: to organize open consultation with representatives of thematic working groups and standardization bodies where projects can identify the challenges, needs, and findings on potential contribution to standards. More confrontation moments on AI Act and its application to the Innovation and Research projects are needed. Where and Who was involved: EC officers. Target audience: RTO, Academies, Large industries, EC officers and agencies, Standardization bodies, working groups. 3.2. TENSOR use cases The TENSOR platform and its technology offerings were evaluated across three different use cases (UCs). The first pilot demonstration and validation phase took place in November 2024. The aim of the first pilot phase was not only to validate the platform, but also to gather valuable feedback from the end-users. This feedback will be the basis for enhancing the technologies and potentially improve TENSOR for the final pilot phase. • The UC1 “Evidence collection through intelligence derived from correlated physiological and behavioural biometrics based on CCTV footage” is a multi-modal and multi-contextual scenario to demonstrate the effectiveness in identifying and verifying the identity of a suspect. The Policejní Prezidium České Republiky (PCR) lead this pilot and validated different aspects of the platform throughout a demonstrator and an on-site workshop. • The UC2 “Digital Forensics Extensions allowing Orphan Device Owner Identification” demonstrates under the leadership of the Ministry of the Interior of Finland (MOI) TENSOR’s digital forensics extension to extract information of interest from an orphan device, i.e., when the owner of a seized device is not known. The extracted information can be fed into an active investigation and might be used to identify the owner of said device. Furthermore, this UC focuses on the integration of novel behavioural biometric technologies of TENSOR into the device owner identification process. • The UC3 “Biometric data protection and secure exchange in a cross-border scenario” demonstrates a data exchange between Law Enforcement Agencies (LEAs) in a cross-border scenario through the European Biometrics Dataspace. This technology has the potential to streamline the data exchange process and by doing so accelerate international investigations. The partners involved in this UC are the Ministério da Justiça (PJ) of Portugal and the Inspectoratul General al Poliției (GPI) of Moldova. To evaluate and validate the TENSOR platform and its components, 60 forensic experts from various LEAs were recruited, who have expert knowledge in the forensic work with biometric data, digital forensics or criminal investigations. Each participant had the possibility to conduct an online training to familiarise themselves with the TENSOR components by completing modules on a Moodle-based website. Any user feedback on the training itself was directly collected via a questionnaire on the Moodle platform. Depending on the UC, the volunteers had then the possibility to gain insights in the TENSOR platform by participating in technology-specific on-site workshops or conducting extensive group testing sessions (in groups of 3) and the ability to interact with the TENSOR platform. Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 23 The workshop and testing sessions were concluded with each participant providing their opinions and insights participating in an online evaluation survey. This enabled us to collect the user feedback during the pilots and identify potential improvement for the second pilot iteration. 3.2.1. Lessons learnt This section summarises the lessons learnt from the TENSOR project, collected throughout the whole piloting phase: from the planning and preparation of the pilot studies, over the training of the participants to familiarise with them with the TENSOR system and its components, and finally during the execution of the distinctive pilots. 3.2.1.1. General feedback The measured operational KPIs focus on the LEAs’ needs to operate the TENSOR platform for biometric data analysis and cross-border biometric data exchange. Some of the defined operational KPIs could already be achieved during the first pilot phase, and with the collected user feedback we aim to achieve the target values of the remaining KPIs in the second iteration of the pilots. In detail, the following KPIs (Figure 11) are assessed during the pilot phases. • O1: Increased efficiency in criminal identification by 60% (achieved) • O2: Improvements of search/classification speed by 25% (achieved) • O3: Improvement of operational standards and processes for LEAs by 70% (in progress) • O4: A percentage increase of automated processes across the investigative workflow by 70% (in progress) • O8: User Acceptance Percentage at >96% (in progress) Figure 11: KPI evaluation results after the 1st pilot iteration Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 24 Feedback on the increased efficiency in criminal identification (O1) stated the need for providing a unified platform to gather and process biometric evidence in an investigation. A good performance of the overall system was stressed as essential in providing fast insights, increasing search and classification speeds (O2). The main reason behind this is the limited human and time resources during the forensic work and the analysis of large volumes of data. End users also highlighted the benefits of processing, fusing, and sharing multiple biometric modalities on one single platform, which will improve existing processes across the investigative workflow (O3), with some of the offered technologies complementing already existing commercial tools. Also, some users remarked that the successful implementation of such a solution depends on the integration of already existing databases and registers within TENSOR, as well as the accompanying legislation. However, according to the end users the integration with legacy systems might not be feasible, as the specifications on the systems and database schemes are considered confidential information and not be shared outside the LEAs. As a direct consequence, this also impacts any interoperability requirements that may not be fully met. A successful integration of a system with existing legacy systems might significantly facilitate the end users’ everyday work, instead of having “yet another platform” that cannot communicate with existing systems and imposing unnecessary user workload. Overall, the end-users recognized the potential of the offered solutions and clearly showed an interest in them. In particular, the ability to fuse different biometric modalities was positively highlighted. Also, they appreciated the potential of the TENSOR solution in enhancing efficiency and increasing automation (O4) compared to methods, technologies and operational workflows within their agency. 3.2.1.2. User training The user feedback regarding the online training encompassed the need to create an interactive and more engaging environment and the need to improve the overall quality of the provided training materials, rather than just providing “user manuals”. End users stated the wish for more hands-on, practical training scenarios. The training should also be tailored to different levels of expertise, preliminary knowledge on the used technologies from potential users, and their specific roles. The different levels of expertise demand for customised training content that links the training material with the practical application of the presented technology. 3.2.1.3. Pilot validation and demonstration Feedback regarding the pilot validation and demonstration itself addressed the need for realistic conditions of the demonstrated platform with the complete set of the presented functions. This encompasses not only the functionalities of the TENSOR components, but also the realism of the used dataset and files. The TENSOR platform should also be tested and deployed on premises during the pilots, which was not possible for the first pilot phase. The needed hardware and connection capabilities at the LEAs’ premises can pose a critical factor for a successful pilot. On the one hand, the end users work in a security critical environment that limits the possibility to connect to any online services. On the other hand, the technical requirements of the technologies in question might surpass the hardware capabilities available at the testing location, creating the need for third parties to be involved in hosting the services. Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 25 3.2.1.4. Usability Usability feedback addressed missing functionalities that may ease the forensic work while interacting with the system. This includes structuring visual components with meaningful information (e.g., chat views for extracted conversations from mobile devices, previewing file contents), the ability to view/interact with various file types (e.g., map for location data, video player), sophisticated search and filter mechanisms to efficiently reduce the number of data, and finally using and exporting standard file formats (e.g., NIST file format for storing biometric data). One suggestion to improve usability is to include the end users in the development process early on, which might aid finding and prioritising the needed functionalities for the platform. At the same time, users praised the user-friendly environment and the ability to initiate the analysis with simple means and the easy creation of reports, though in-house standardised procedures must be considered to ensure that evidence can be used in court. 3.2.1.5. Technological aspects AI poses a potential technology for aiding forensic experts in their everyday work for finding biometric matches and the identification of persons. Nevertheless, many AI-based models have their own definitions for similarity scores and thus making the interpretation of the results, e.g., likelihood or the level of confidence, very difficult, which might be needed as a reasoning to defend the results in court. Thus, information on the model’s internal mechanisms might be a crucial addition. Independent from the type of technology, the end users demanded for each technology to include an explanation on the following points: • its capabilities (what it can and cannot do) • its functionalities (how to use the technology) • understanding the technology (how does it create results) • interpreting the results (what does this result mean) This information might not just be part of the training but also be integrated in the final platform to provide context-based on-demand help for users. 3.2.1.6. Legal aspects Although inside the EU, the member states are bound to the GDPR, we found that the individual partners from different countries embrace different levels of national laws, that expand on the GDPR. Apart from national laws, the LEAs might impose additional in-house policies that makes the processing and sharing biometric data between countries particularly challenging. Also, using real datasets from criminal records was not possible to validate the components as they are not available or restricted by law and thus posing a special challenge for creating a realistic testing environment. Therefore, we needed to consider alternatives to obtain realistic data (e.g., from volunteers) and enrich the datasets to a realistic volume (e.g., using synthetic data). Any legal discussions towards implementing data processing agreements or consent forms should therefore be initiated as early as possible in the project. Based on our past experiences these processes tend to be tedious and lengthy, as various (legal) departments of the affected partners need to be included in the discussions. Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 32 Recommendation 1: Prioritise the adoption of existing international standards for biometric data interchange, such as the ISO/IEC 19794 series, which defines specifications for various biometric modalities (fingerprint, face, voice). Ensuring compatibility with existing frameworks for cross-border law enforcement cooperation and data exchange will further support interoperability. Strategy: Actively participate in standardisation bodies where possible and contribute to refining and extending standards to address forensic application needs. Focus on developing standardised APIs and data formats for biometric modalities, enabling seamless integration with legacy systems and avoiding vendor lock-in. Harmonising data formats and protocols will facilitate efficient data exchange between biometric platforms and existing law enforcement infrastructures. 3.2.3.2. Data Sharing and Governance A harmonised approach to biometric data sharing will enable LEAs to collaborate securely while ensuring adherence to legal and ethical standards. Establishing a trusted data exchange ecosystem will enhance operational efficiency and promote responsible handling of sensitive biometric information. Recommendation 2: Establish a standardised framework for data sharing agreements between LEAs, incorporating clear guidelines on data usage, access control, and liability. The framework should align with international best practices and relevant legal frameworks to ensure lawful and ethical data exchange. Strategy: Develop reusable templates for data sharing agreements, covering key aspects such as data ownership, purpose limitation, and security requirements. Introduce a certification process for data providers to ensure compliance with these agreements. The framework should also adhere to fundamental privacy principles, such as transparency, accuracy, and storage limitation, to maintain trust and legal compliance. 3.2.3.3. Privacy and Security Privacy and security concerns surrounding biometric data are critical, and standardised privacy-enhancing technologies will provide LEAs with secure data processing methods that do not compromise individual privacy. Strengthening security measures, such as robust encryption, access controls, and authentication mechanisms, will help build public trust in biometric technologies. Recommendation 3: Advocate for the adoption of standardised privacy-enhancing technologies for biometric data processing, such as homomorphic encryption and secure multi-party computation. Implementing these technologies will help minimise data breach risks and enhance compliance with privacy regulations. Strategy: Conduct research on the performance and scalability of privacy-enhancing technologies in biometric applications. Develop reference implementations and encourage adoption through open-source initiatives and industry collaborations. Ensure data minimisation techniques are in place, aligning processing activities with their intended purpose to maintain privacy safeguards. 3.2.3.4. Ethical AI Ethical AI practices are fundamental to mitigating bias and ensuring fair outcomes in biometric applications. Standardisation in this area will help define clear accountability measures and ensure that biometric technologies are deployed responsibly, reducing the risk of discriminatory outcomes. Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 33 Recommendation 4: Promote the development and adoption of ethical guidelines, benchmarks and standards for AI in biometric applications, ensuring transparency, fairness, and accountability. Addressing bias and discrimination risks is essential for responsible AI use in law enforcement, and standards around metrics and bias auditing could contribute to the development of responsible AI practices. Strategy: Develop best practices for responsible AI development, incorporating principles of explainable AI (XAI) and fairness-aware machine learning. Promote an AI auditing methodology to help developers understand and incorporate regulatory expectations related to human oversight bias, explainability and risk mitigation generally. Establish a certification process for AI systems to ensure compliance with ethical standards. AI systems should be designed to be unbiased and non-discriminatory, fostering trust, transparency and fairness in their application. 3.2.3.5. Digital Forensics and Device Security Standardisation in digital forensics is essential to ensure the integrity and admissibility of biometric evidence in investigations. Creating awareness throughout all participants and establishing clear guidelines and methodologies will enhance the effectiveness of digital forensic practices and support lawful access to critical biometric data. Recommendation 5: Develop standardised guidelines for digital forensics investigations involving biometric data, focusing on secure device unlocking, data extraction, and chain of custody procedures. These guidelines should align with international best practices for handling digital evidence. Strategy: Collaborate with LEAs and forensic experts to establish practical methodologies for digital forensics investigations. Promote the use of secure, standardised tools and techniques for data extraction and analysis to maintain forensic integrity. By actively engaging in standardisation efforts and promoting the adoption of these recommendations, biometric technology initiatives can contribute to a more secure, interoperable, and ethically responsible ecosystem for law enforcement applications. This will not only enhance operational efficiency but also strengthen public confidence in the use of biometric technologies. 3.3. ODYSSEUS use cases 3.3.1. Lessons learnt The ODYSSEUS project aims at validating and demonstrating eight key technical solutions through five distinct pilots (2 Train Pilots, 2 Land Pilots and the Water Pilot), implementing a diversity of Use Case Scenarios. The main objectives of this phase were to develop a comprehensive evaluation and validation plan, engage end-users for pilot demonstrations, assess ODYSSEUS platform performance, validate ODYSSEUS solutions across different environments, and analyse the pilots’ results for further improvements, ensuring the integral execution of them, and aligning with ODYSSEUS key performance indicators (KPIs). To this end, a structured methodology, as depicted in the following figure (Figure 12), was designed and established in order to bridge the gap between the initial project objectives and the real-world constraints encountered during the pilot phase, linking each objective with an ‘action’ phase adhering to a timeline. Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 34 Figure 12: Pilot preparation and demonstration methodology Our concept consisted of five consecutive phases, starting from the most critical, the PilotDefi - nition, where our major efforts were devoted to identify the pilots’ limitations, the technology providers’ pilot specific requirements, and the technology selection to be implemented in each pilot. The two lessons learnt from this phase are presented below. Lesson 1: Technologies’ Comprehension from Stakeholders’ View is Critical for Technology Selection and Implementation: Stakeholders’ understanding of the technological solution that are going to be deployed is crucial for decision-making in the context of technology selection and effective implementation. In particular, in the case that the stakeholders, especially end-users, do not fully grasp how a solution works or what it entails, this can result in misalignments, mismanagement, and underperformance during the pilot execution. Lesson 2: Pilots’ Limitations/Regulations and Technology Deployment Requirements Definition are Crucial for Smooth Pilot Execution It is of high importance to clearly understand the constraints and the regulations that a pilot has to compliant, and also, to define the requirements for technology deployment. In some cases, photos or indicative videos of the pilot facilities should be provided for a better view and comprehension, or even more an onsite visit could be a good idea. Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 35 Mitigation measures: Technology providers must clearly communicate the functionalities, limitations, and potential benefits of their solutions. This communication should not only emphasize on technical specifications, but also, should make sure that the end-users can perceive how the technology will be utilised and what might be impact on their processes. The second phase, Planning, incorporated the pilots’ availability and, respectively, the logistical barriers per technological solution. During this phase, we allocated the Use Case Scenarios per pilot along with the stakeholders’ definition, and we interpreted the former to demonstration/execution Scenarios. The lessons learnt consist of: Lesson 3: Considering availability and logistical barriers prevents delays Failure to take into consideration the logistical challenges, such as equipment transport and infrastructure readiness, can lead to significant project delays. Lesson 4: Translating scenarios into operational demonstration plans enhances end-users’ and tech providers’ understanding Simply defining use case scenarios is not enough; translating them into actionable, step-by-step demonstration plans ensures that all stakeholders, and especially end-users, grasp the technical aspects and expected outcomes. This enhances collaboration and reduces misalignment between expectations and implementation. Lesson 5: Defining and engaging stakeholders improves acceptance and increases the likelihood of smooth pilot execution Engaging key stakeholders from the outset ensures their needs and constraints are addressed, fostering greater acceptance of the technology. Lack of early involvement can result in resistance, operational conflicts, or unanticipated integration challenges. Mitigation measures: Close cooperation between end-users and tech providers ensures that operational requirements and constraints are continuously addressed, preventing misalignment and ensuring smoother execution. Afterwards, the Implementation phase followed, when some of the selected solutions were deployed to the pilot sites beforehand, because of logistical and commissioning constraints, Additionally, we introduced the technological solutions outlining the Use Cases Scenarios to the end-users, and simultaneously, we were presenting the functionalities of the ODYSSEUS platform as a training workshop. Lessons learnt 6 and 7 are representative of this phase. Lesson 6: Comprehensive testing of systems across diverse environments requires adjustments to scenarios and components, preventing integration failures A technological component working in one specific environment may not function or have to be adjusted in another, due to variations in infrastructure, connectivity, or environmental conditions. Iterative testing across different pilot sites allows for scenario refinement and transformation to operational ones, and for system optimization. Lesson 7: Organizing training sessions and hands-on workshops prior to execution is crucial End-users must be adequately trained before deployment to ensure they can operate the system Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 36 effectively. Hands-on workshops help ensure that end-users are fully equipped to interact with the new system, facilitating smoother adoption. Mitigation measures: 1. Tech partners adjust their solutions according to specific scenarios, incorporating buffer time for testing and troubleshooting. 2. Briefing and debriefing sessions facilitate knowledge transfer and end-user adoption. The fourth phase of our plan, titled Demonstration, is strongly related to the previous ones, though we have to distinguish them, due to the latter acts as the actual demonstration process with the physical presence of all relevant partners and pilot execution in real world conditions. The lessons learnt identified during this phase are introduced as follows. Lesson 8: Real-world testing and active stakeholder engagement provide critical insights into system performance and offer feedback for continuous improvement Pilots’ execution in real-world environment help reveal unpredictable issues. Feedback loops from stakeholders can provide fine-tuned aspects for the ODYSSEUS integrated solution. Lesson 9: External factors, such as weather conditions, transport delays (train/ferry cancellations), or long vehicle queues, should be taken into serious consideration Unforeseen external conditions can derail pilots’ execution plans. By incorporating flexibility into scheduling may mitigate such disruptions. Mitigation measures: 1. Extensive discussions with stakeholders gather insights into potential disruptions and necessary adaptations. 2. Flexible scheduling and back-up plans help adjust to real-time challenges, ensuring smooth operations. Last but not least was the Evaluation phase which is highly pertained to the Pilot Validation, Pilot Assessment and User Acceptance Evaluation, and its title reveals that it assesses the effectiveness of the demonstration through performance metrics and KPIs. The lessons learnt are following. Lesson 10: User experience and ethics questionnaires/surveys preparation raises awareness and transparency Diving into user experience and ethical considerations is critical for technology adoption and helps raise awareness among stakeholders and ensures transparency in the pilot processes. These surveys also provide valuable data on user perceptions and system performance, improving acceptance, identify concerns, and ensure compliance with ethical regulations. Lesson 11: Assigning KPIs per partner and clearly defining measurement methods enhances accountability and performance tracking Assigning KPIs to partner ensures accountability and transparency, resulting in aligning expectations on the project’s outcomes. Mitigation measures: 1. Expert-prepared, GDPR-compliant questionnaires ensure transparency and regulatory compliance. 2. Early KPI allocation, along with a well-structured responsibility matrix, ensures effective monitoring and reporting. Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 37 As a key takeaway of the ODYSSEUS project, the critical role of the end-users’ thorough understanding in the technological solutions to be developed. This might help in a wider adoption, which hinges on clear communication, comprehensive training, and iterative feedback from them. It is notable that the project demonstrates that a collaborative, user-centred approach enhances the viability of ODYSSEUS innovations and maximizes the long-term ODYSSEUS real-world impact. 3.3.2. Gap Analysis In ODYSSEUS the gap analysis was conducted at two axes: • Analysis of state-of-the-art technologies and processes was conducted by workshops with border control partners and technological experts. • Surveys were conducted for travellers, border control officers and technological border control experts to collect perception of current processes and define requirements of the ODYSSEUS platform as perceived by the stakeholders. The results of the border control state of the art provided following improvement points: • Cost of the border crossing: cost of deployment of new equipment and training of new staff is a significant problem for the BCA and reduction in either would improve both the efficiency of the process and traveller experience. • ExpertiseofBCAstaffandprocessguidance: the efficiency of the process and ability to detect fraud strongly depends on experience of the BCA teams. The experience impacts the ability to detect document forgery, identification of suspicious behaviour, ability to find illicit goods and human trafficking. Deployment of guidance systems that would increase detection rates for less experienced staff would be an improvement to security of the border crossing. • Deployment of advanced tools: even though there are advanced tools and technologies available on the market they are rarely deployed due to cost reasons. In review of technologies selected for border crossing in ODYSSEUS we identified following gaps: • Paperless border crossing: usage of digital credentials for travel are at early stage of deployment. Even though the specification has been ready for several years there is no large-scale deployment, only pilots are ongoing and all activities are related to border crossing for airports. The deployment models have several gaps: there are legal and organizational gaps in who will be responsible for identification of the travellers, who will be responsible for issuance of the DTC and who will be responsible for the data processing of the travellers. At technical levels we identified a few gaps that prevent deployment of the DTC on smartphones when it is currently impossible for iOS and deployment on Android is possible only with a workaround. • Touchless border control based on face match technologies is currently in pilot mode for airport border crossing but was not yet tested or piloted in other types of crossing. • UAV deployment for border crossing is hindered by technological gaps in the UAVs themselves. The UAVs for deployment in such conditions need high battery time, high load due to additional sensors needed to collect required information, they also need to be able to transmit high amount of data collected from the sensors. Process wise the UAVs need to be autonomous during the scanning to reduce staff needed for their deployment. Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 38 • X-ray scanning technology currently has low dosage, but still too high enough to be regulated for large-scale deployment. Current solutions are focused on container scanning for illicit goods, but in BC environment the detection of unauthorized persons is as important. There is a lack of solutions capable of detection of chemical, biological and explosive threats. Current solutions may be improved by AI usage to increase detection for less experienced staff. • Crowd monitoring and counting was never deployed in BC environment before and parameters of such solution such as crow density detection, occlusion need to be adjusted and the solution deployment customized for BC lighting conditions and camera placement. We also expect the crowd dynamics will be different than in standard conditions the solution was deployed in. • AI and XAI deployment for decision support in BC conditions is an ODYSSEUS innovation and we expect several gaps to be addressed:  Accuracy: such solution needs to provide accurate decision support which will rely on cor - rectness and representativeness of training data.  Security and privacy: the solution will be operated in highly sensitive environment and handle personal information. Explainability: the decision support system needs to provide optimal guidance and transparency regarding the risk assessment. The system may have impact on fundamental human rights, so it needs to provide maximum information why the traveller risk was assessed. Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 39 3.3.3. Recommendations on possible standardization activities and strategies for policy experts 3.3.3.1. Overview The ODYSSEUS project aims to enhance border security and operational processes through a well-structured standardisation strategy. The project is divided into three phases. In Phase 1, the Standardisation Task Force is established, and preliminary mapping is conducted through an online survey to gather initial data and insights. Phase 2 involves engaging project partners and end-users in workshops to ensure collaborative development. During this phase, a detailed analysis of innovations is performed, and the Action Plan is defined. Phase 3 focuses on implementing the Action Plan, with continuous monitoring, evaluation, and reporting in last deliverables of the project. 3.3.3.2. Applicable Standards Identified Several applicable standards have been identified for the ODYSSEUS project. For digital and virtual passports, the ICAO Digital Travel Credentials (DTC) and ICAO Doc 9303 are relevant. UAV-assisted X-Ray technologies should comply with ISO/IEC 27001, ISO 9001:2015, and ISO 14001:2015. Seamless identity verification in border crossing scenarios must adhere to ICAO standards and ISO/IEC 19794-5 compliance. Multi-modal fusion and AI-based decision support systems should follow ISO/IEC 23894:2023, ISO/IEC TR 24028:2020, and ISO/IEC TR 24027:2021. Behavioural authentication models need to comply with ISO/IEC 27001, ISO/IEC 23894:2023, and ISO/IEC TR 24028:2020. 3.3.3.3. Feedback and Contributions Feedback and contributions are essential for refining these innovations: • Digital and Virtual Passports: Usability on mobile devices, Integration with legacy systems. • UAV-assisted X-Ray Technologies: Safety standards, Ethical considerations, Data protection standards, Interoperability, Regulatory compliance, Quality assurance, Environmental impact. • SeamlessIdentityVerification: Efficiency standards, Cost reduction standards, Quality standards. • Multi-modal Fusion and AI-based Decision Support Systems: Quality decision-making standards, Trustworthy AI principles. • Behavioural Authentication Models: • Anonymization, Encryption, Ethical AI principles. 3.3.3.4. Engagement with Standardisation Bodies Engagement with relevant standardisation bodies is a key part of the strategy. The ODYSSEUS project will be presented to the ICAO NTWG on the topic of future forms of eMRTD and associated systems. Additionally, the project will be showcased at the CEN TC 224 Plenary meeting in June 2024. These presentations will help position the ODYSSEUS project as a leader in border security innovations and standardisation. Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 40 3.3.3.5. European Commission’s Standardisation Strategy Launched on February 2, 2022, it aims to: • Bolster EU Leadership: Enhance the EU’s leadership in global standards. • High-Level Forum: Establish a High-Level Forum on European Standardisation to set priorities. • Improve Governance: Enhance governance within the European standardisation system. • Strengthen International Standardisation: Strengthen the European approach to international standardisation. • Support Innovation: Connect research and innovation with standards. • Promote Academic Awareness: Increase academic awareness and prepare future standardisation experts. 3.3.3.6. Conclusion • In conclusion, the ODYSSEUS project’s standardisation strategy is designed to ensure that innovations are efficiently delivered. By engaging with key standardisation bodies and adhering to identified standards, the project aims to enhance border security and operational efficiency. 3.4. TENACITy use cases The TENACITy project aims to enhance travel intelligence and security through the implementation of advanced data analysis, risk management, and interoperability frameworks. The project’s overarching objective is to modernize travel intelligence governance through the following key objectives: • Increasing Data Reliability by enhancing the accuracy and consistency of PNR data through advanced analysis and pattern identification tools. • Strengthening Risk Management through cutting-edge AI and agent-based modelling approaches to predict and mitigate security threats. • Enabling Secure Data Exchange through blockchain technology to safeguard data integrity and privacy. • Promoting Interoperability by ensuring seamless integration of new tools into existing operational environments. • Enhancing Stakeholder Trust by applying societal frameworks to improve transparency and citizen engagement. While each case study is tailored to its specific operational environment, they share these common goals. Use Case “Known Suspect Traveling Between Member States” This piloting case study addresses the challenge of tracking a known criminal moving between European countries. Due to the absence of API data for intra-Schengen movements, the use case emphasizes the improvement of PNR data reliability by collecting and analysing data from multiple points along the suspect’s journey. By leveraging risk management and pattern identification, law enforcement agencies from Greece and Spain will collaborate to increase situational awareness and exchange reliable travel intelligence. Key innovations include: Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 41 • Advanced PNR data verification and enhancement through cross-border collaboration. • Real-time behavioral pattern analysis to track suspect movements. • Secure data exchange through blockchain for reliable and tamper-proof information sharing. UseCase“FirearmsTrafficking” This piloting case study targets the detection and prevention of illegal firearms trafficking into the EU via third countries. Utilizing open-source intelligence from darknet data sources, the project aims to track operational patterns of firearm trafficking networks. The solution will integrate resilience-oriented risk management, addressing systemic risks associated with violence enablers. By combining OSINT with travel intelligence governance, law enforcement and customs authorities will collaborate to mitigate firearm smuggling threats. Key innovations include: • Advanced threat pattern detection using OSINT and AI-driven risk assessment. • Secure cross-border data exchange for coordinated response. • Real-time tracking of criminal networks via advanced pattern identification techniques. Use Case “Lone Terrorist or Small Group Entering Europe” This case study focuses on the challenge of identifying unpredictable movement patterns of lone terrorists or small groups entering Europe from third countries. By employing advanced pattern identification combining AI with agent-based modelling, the system aims to detect emerging risks from individual or small group behaviors, predicting their next movements and possible threats. Collaboration between UK and European LEAs will ensure comprehensive data integration and rapid response to emerging threats. Key innovations include: • Combining AI-driven pattern recognition with agent-based modelling for enhanced threat prediction. • Real-time data analysis from diverse data sources to identify and mitigate risks. • Secure and transparent data sharing through blockchain technologies. Use Case “Virtual Setting” In this unique piloting case study, a virtual simulation environment will be created to test interoperability and integration of TENACITy’s solutions with existing law enforcement frameworks. The primary goal is to demonstrate the project’s capability to enhance collaboration among diverse stakeholders while preserving data security and privacy through blockchain technologies. Key innovations include: • Virtual integration of multiple data sources and operational frameworks. • Comprehensive testing of interoperability between new and legacy systems. • Privacy-preserving data exchange with blockchain. Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 48 4. Conclusions In this section the four projects have summarized the common needs and the commonalities that emerged during the different iterations and the two synergy workshops that have been organized. 4.1. Commonalities The projects highlighted the importance of systems interoperability even if each project is using a different approach to exchange data because of the context. Biometric data exchange among LEAs, people travel experience (notwithstanding the fragmented travel intelligence governance) and data interoperability among BCP are of primarily importance. But the interoperability suffers of two constraints: legal and regulatory constraints, the confidentiality and the lack of real data. Moreover, two other commonalities have been identified: • the usage of the blockchain for accessing people/travellers and goods information • the extraction of patterns from raw unstructured data 4.2. Needs 1. Common Methodology for Risk Assessment: to establish a standardized risk assessment methodology that can be applied across all operations, processes, or security projects. This will ensure consistency and allows for the integration of diverse risk factors (e.g., financial, technical, operational) defining a unified data model whose output should be pursued by the projects. 2. Training: to recommend the dissemination leaders to implement training programs during the projects focused the usage with proficiency of the developed tools. Training should be planned with a market-oriented profile, in order to have the end users of the projects as the primary customers of the tools. 3. Uniform Method for Data Generation: to develop a consistent method for generating synthetic data to be used as a knowledge base for the different topics of HORIZON clusters. This could involve defining clear data standards (formats, structures, sources) to ensure that all data is comparable, accurate, and easy to analyse. Data management platforms or frameworks, such as ETL (Extract, Transform, Load) processes, can help standardize data workflows, ensuring uniformity and quality. All projects could start from this knowledge base enriching it at the end of the development and testing cycle. 4. Best Practice for Extracting Operational Scenarios and requirements: Extracting operational scenarios requires a structured approach to ensure that all relevant information is captured and the scenarios are useful for decision-making, forecasting, and tactical planning. The need here is to build a knowledge base that should contain all the historical data from which to extract recurring operational patterns. The knowledge base could be built by leveraging automation for data collection and scenario generation. This could include data analytics platforms, AI models, or specialized software to collect, analyse, and Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 49 generate operational scenarios in real-time, reducing manual errors and saving time and above not disclosing to requestors of new scenarios all the confidential information to similar scenarios stored by different project on behalf of different end users participating to the projects. The knowledge base could be useful also to store requirements and to extract patterns from the knowledge base through the usage of AI tools in order to suggest very well consolidated and standardized requirements The implementation of these needs will would streamline processes, reduce inconsistencies, and ensure that all stakeholders of research projects are aligned in terms of risk management, data generation, and scenario extraction and training. Strengthening Security Through EU-funded Research & Innovation: Advancing Border and Forensic Capabilities Joint Report, August 2025 50 5. Acknowledgements FLEXI-cross – Funded by the European Union under Grant Agreement no. 101073879. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Research Executive Agency. Neither the European Union nor the granting authority can be held responsible for them. TENSOR – Co-funded by the European Union under Grant Agreement no. 101073920. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Research Executive Agency. Neither the European Union nor the granting authority can be held responsible for them. ODYSSEUS – Unobtrusive Technologies for Secure and Seamless Border Crossing for Travel Facilitation. ODYSSEUS has received funding from European Union’s Horizon Europe Innovation Programme under Grant Agreement N°101073910. Content reflects only the authors’ view and European Commission is not responsible for any use that may be made of the information it contains. TENACITy – Funded by the European Union under Grant Agreement no. 101074048. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or REA. Neither the European Union nor the granting authority can be held responsible for them.