Open Science in transport research: insights from the Maritime and CCAM pilot cases

12ο ΔΙΕΘΝΕΣ ΣΥΝΕΔΡΙΟ ΓΙΑ ΤΗΝ ΕΡΕΥΝΑ ΣΤΙΣ ΜΕΤΑΦΟΡΕΣ Οι μεταφορές στην εποχή της Τεχνητής Νοημοσύνης 12th INTERNATIONAL CONGRESS ON TRANSPORTATION RESEARCH Transportation in the era of Artificial Intelligence Open Science in transport research: insights from the Maritime and CCAM pilot cases 1 Open Science in transport research: insights from the Maritime and CCAM pilot cases Afroditi Anagnostopoulou1*, Athanasios Ballis2, Xenophon Kitsios1, Anastasia Bolovinou2, Vassilis Kappatos1, Angelos Amditis2 1. Centre for Research and Technology Hellas / Hellenic Institute of Transport, Greece 2. Institute of Communication & Computer Systems, Greece *a.anagnostopoulo[email protected] Abstract As transport research is increasingly data-driven, the implementation of FAIR (Findable, Accessible, Interoperable, Reusable) principles and open science practices is crucial to foster innovation, transparency and collaboration among stakeholders. This paper presents the results of two pilot cases in the SCILAKE project, Maritime Transport and Cooperative, Connected and Automated Mobility (CCAM), exploring real-world applications of the FAIR data principles. The maritime pilot addresses the challenges of interoperability and data access and aims to improve operational efficiency and situational awareness through digital platforms and open standards. The CCAM pilot project focuses on secure and equitable data sharing for automated transportation systems in different environments and promotes better collaboration between academia, industry and public stakeholders. In both cases, important factors such as governance models, regulatory harmonization, artificial intelligence and knowledge graphs are highlighted. The study provides practical recommendations for the implementation of FAIR data in transportation and emphasizes the importance of ethical considerations and multi-stakeholder engagement to support a transparent, inclusive and sustainable digital transformation in the sector. Keywords: Open Science, FAIR data principles, Data governance, Transport research. 1. Introduction Transport research increasingly relies on digital data, so the application of Open Science principles is critical to fostering innovation, reproducibility and stakeholder trust. Open Science plays a key role in transport research by promoting transparency, collaboration and real-world impact. It increases the credibility and reproducibility of research through open access to data, methods and results, and enables stakeholders to review and build on results. In addition, the FAIR (Findable, Accessible, Interoperable, and Reusable) principles provide a structured approach to data management that improves transparency and collaboration. More specifically, Open Science accelerates knowledge sharing between academia, industry and government and promotes cross-sector collaboration and innovation (Reichmann and Wieser, 2022). It also supports evidence-based policymaking and ensures that publicly funded research maximises societal benefits (Olesk et al., 2019). By breaking down institutional silos, Open Science helps to address complex mobility challenges with more integrated and sustainable solutions. As transport systems Open Science in transport research: insights from the Maritime and CCAM pilot cases 2 become increasingly data-driven and connected, the openness of scientific results becomes essential for harmonising standards, aligning research agendas and ensuring broad access to innovation. Ultimately, Open Science enables a more responsive and equitable approach to shaping the future of mobility. Ensuring FAIR data in transport research is critical to maximising the value of data for all stakeholders. It enables the efficient reuse of existing datasets, reduces redundancies and lowers research costs (Martorana et al., 2022). FAIR data facilitates collaboration between academia, industry and government by providing a common basis for data integration and analysis. It also supports evidence-based policy making, promotes innovation in emerging mobility services and improves transparency and accountability (Langbridge et al., 2024). Importantly, FAIR data contributes to the development of more sustainable, resilient and interoperable transport systems, in line with the broader goals of digital transformation and environmental protection. However, implementing these principles in transport research is challenging due to legal requirements, data ownership concerns and technical barriers. This paper presents findings from two pilot projects within the SCILAKE project: (i) Maritime Transport and (ii) Cooperative, Connected and Automated Mobility (CCAM). Based on these pilots, we analyse how Open Science can be implemented in practical cases and identify best practices, barriers and enablers for the implementation of FAIR data. Both pilots serve as a test environment for the application of the SCILAKE tools to real challenges in the transport sector and provide information on how FAIR-oriented digital infrastructures can support transparent, secure and interoperable data exchange. By examining the respective regulatory, technical and operational contexts, the paper shows how sector-specific adaptations of Open Science principles can lead to systemic improvements in transport research. The remainder of the paper is organised as follows: Section 2 analyses how Open Science and FAIR data approaches are applied in transport research and outlines why this is important. Section 3 presents and discusses the two pilot cases of maritime transport and CCAM. Section 4 then summarises the transport research insights and best practices for enabling Open Science in this sector. Finally, Section 5 draws conclusions and makes recommendations that emphasise the importance of multi-stakeholder collaboration and regulatory alignment for the promotion of Open Science and FAIR data in transport research. 2. Open Science and FAIR data in transport research Open Science (Vicente-Saez and Martinez-Fuentes, 2018) is a movement that promotes transparent and accessible research processes and ensures that data and methods are shared between different stakeholders. Applying FAIR principles (Mons et al., 2020) to transport research improves data quality, usability and integration across different systems. However, transport datasets are often complex and involve multiple stakeholders, legal requirements and technical standards that limit open access and interoperability. Open Science therefore includes practices such as open access publishing, data sharing and collaborative platforms for joint development in transport research. The FAIR principles, first formulated by Wilkinson et al. (2016), are increasingly recognised as fundamental to data management strategies. These principles ensure that data can be discovered, retrieved under clear conditions, integrated with other datasets and reused with appropriate documentation. The importance of FAIR data in transport-related areas was recently highlighted in the study by Bokolo and Sarshar (2025), who point out that the effective implementation of FAIR data in mobility research improves interoperability between systems and actors, especially in multimodal transport. Kitchin (2023) also argues that open data in smart cities must be embedded in the governance framework in order not to reinforce existing power asymmetries. To this end, Sabouni et al. (2023) note a growing trend for public authorities to adopt FAIR-oriented data strategies to improve traffic management, environmental monitoring and infrastructure planning. Open Science in transport research: insights from the Maritime and CCAM pilot cases 3 Despite the developments mentioned above, there are still challenges. According to Janssen et al. (2012), the openness of data in transport is often hindered by institutional silos, lack of standardisation and unclear ownership of data. Furthermore, Hollnagel and Fook (2019) warn that open data initiatives can stall without proper stakeholder coordination and technical infrastructure. Recent European initiatives such as the European Commission's Common European Mobility Data Space (Doulkeridis et al., 2024) and the Data Governance Act (Ruohonen and Mickelsson, 2023) reflect a shift towards the formalisation of Open Science practices in the transport sector. These measures aim to harmonise data access and reuse across European Member States and create a legal and technical framework for the FAIR implementation. In contrast to land transport, data in maritime transport is often linked to cross-border logistics, international regulations and proprietary operational systems. Fernando et al. (2024) examine the digital transformation of the maritime sector and emphasise the role of data standardisation in achieving interoperability between port authorities, shipping companies and customs authorities. They emphasise that the FAIR principles are essential for unlocking efficiencies in optimising port calls and real-time logistics. Recent EU efforts such as the European Maritime Single Window Environment – EMSWe (EMSA, 2023) and the Digital Transport and Logistics Forum – DTLF (2020) also aim to harmonise digital data flows and ensure cross-border interoperability through common standards and open data platforms. Ferreira et al (2017) mention that although maritime navigation systems have started to integrate digital data streams, the fragmented nature of data ownership remains a major obstacle. It is evident that the maritime sector has unique characteristics that make it difficult to apply the principles of Open Science and FAIR Data. Wang et al. (2019) examine how open access to Automatic Identification System (AIS) data has facilitated maritime research and innovation, but also raises concerns about data privacy, quality and misuse. They propose governance mechanisms for responsible data sharing. In addition, Potamos et al. (2024) emphasise the cybersecurity challenges of open maritime data and recommend robust encryption, authentication and compliance measures to support secure FAIR data infrastructures. In addition, the interconnected nature of CCAM systems requires robust data management and openness to ensure secure, efficient and inclusive mobility solutions. Wheeder and Rout (2023) describe the critical need for open data architectures in the vehicle-to-vehicle and vehicle-to-infrastructure domains to enable real-time communication between vehicles, infrastructure and control centres. They argue that FAIR principles could be the basis for building trust and interoperability between platforms. Alonso et al. (2018) emphasise that transparent data management is key to enabling accountability, trust and real-time collaboration between connected vehicles, infrastructure and authorities. Data sharing, supported by clear standards, metadata and stakeholder access, is also crucial for the equitable deployment of intelligent transport systems. Banerjee et al. (2023) particularly emphasise how transparent data sharing accelerates innovation and supports adaptive traffic management by analysing the role of Open Science in Mobility-as-a-Service (MaaS) ecosystems. More recently, Eshetu et al. (2024) examine the importance of traceability of data streams and focus on how decisions made by automated systems can be reviewed and understood by stakeholders. They also discuss in favour of transparency through design approaches that integrate explainability and accountability mechanisms into the architecture of CCAM technologies from the outset. Bolovinou et al. (2024) emphasise the importance of data sharing and fairness in CCAM by presenting a modular, service-oriented architecture that supports interoperability and openness between stakeholders. They present that fair data access is crucial to enable collaboration between different Open Science in transport research: insights from the Maritime and CCAM pilot cases 4 actors, such as vehicle manufacturers, infrastructure providers and service operators, while ensuring accountability and innovation. The architecture promotes transparency by defining clear interfaces and rules for data ownership, which are essential for the trust and scalability of CCAM systems. This is also the direction of the European Commission's Data Act (EU Data Act, 2025), which reinforces the pursuit of open, standardised and ethically regulated data environments in CCAM. It aims to harmonise data sharing practices with the broader goals of digital sovereignty, security and market innovation. To this end, the C-Roads Platform project promotes the application of standards to harmonise C-ITS services and boost interoperability across Europe, relying on common protocols, regulatory decisions and data structure, but this process is not yet standardized and available to any interested party. In the context of Software-Defined Vehicles (SDVs) and their integration with Internet/Web of things (IoT/WoT) services, the data format is a crucial aspect of enabling seamless communication and data exchange, as acknowledged by Connected Vehicle Systems Alliance (COVESA white paper) and original equipment manufacturers (OEMs) (CLEPA position paper). SDVs, with their focus on software-defined functionalities and over-the-air updates, generate vast amounts of data that can be leveraged for various IoT/WoT applications, ranging from vehicle diagnostics and predictive maintenance to personalized user experiences and vehicle-to-everything (V2X) communication. Structured data is also promoted for safety-critical services, like the European Next Generation Emergency Call that is already deployed in many EU countries, containing standardized content and format promoting a minimum set of data (Sdongos et al 2016). The data collection of events and incidents on the function of autonomous vehicles is helpful in exploring the causes and analyzing the circumstances, thus open structured data is necessary to be provided to the community, as the (AVSC SAE 2020) best practices imply. The US National Highway Traffic Safety Administration (NHTSA) issued a directive requiring manufacturers and fleet operators to report certain crashes involving either advanced driver assistance or higher-level “automated driving systems” (NHTSA SGO), while general traffic accident datasets like the US open Automated Driving Crashes Dataset can be adopted to also include incidents of Automated Vehicles (AV crashes). The literature emphasises the growing importance of Open Science and FAIR data principles in transport research, particularly in improving transparency, interoperability and collaboration between stakeholders. The benefits are obvious and range from improved traffic management and real-time logistics to increased trust and innovation. However, implementation is often hampered by legal, institutional and technical barriers. The maritime and CCAM sectors present unique challenges, such as fragmented data ownership and cybersecurity risks, but also show how targeted frameworks and governance models can operationalise FAIR data in practice. In both areas, data transparency, fairness and clearly defined protocols for sharing are seen as essential prerequisites for scalable, accountable and inclusive transport systems. 3. Pilot Cases 3.1 Maritime transport The maritime transport sector plays a central role in global logistics and yet is one of the most complex and highly regulated areas in the transport ecosystem. The sector is subject to a multi-layered framework of international conventions, regional guidelines and national policies and must fulfil both operational requirements and compliance obligations. Against this backdrop, the SCILAKE pilot project for maritime transport was developed to explore how Open Science practices and FAIR data principles can be applied to improve data sharing, transparency and interoperability of systems. In the past, maritime data has been isolated by different actors such as port authorities, shipping companies, logistics providers, public authorities and researchers, each working with their own digital Open Science in transport research: insights from the Maritime and CCAM pilot cases 5 infrastructure. This fragmentation has hindered real-time collaboration and reduced the overall efficiency of maritime operations. Regulatory instruments such as the International Maritime Organisation’s (IMO) e-Navigation Strategy and the European Maritime Single Window Environment (EMSWe) reflect ongoing efforts to harmonise digital reporting and standardise data exchange across borders. At the same time, regulations such as the General Data Protection Regulation – GDPR (European Commission, 2016) are introducing strict controls on the handling of personal and sensitive data, requiring a balance between openness and robust privacy protection. The pilot project focuses on overcoming these challenges by testing how the FAIR data principles can be implemented in a maritime context. A key objective is to improve interoperability between different data systems and ensure that stakeholders can share information securely and efficiently. This includes not only the adoption of common data formats and protocols, but also the development of a framework that respects legal restrictions while promoting open collaboration. The pilot project thus serves as both a test environment and a reference point for broader digital transformation efforts in the maritime sector. 3.2 Cooperative, Connected, and Automated Mobility Cooperative, Connected and Automated Mobility (CCAM) is a dynamically growing transport sector, including among others key technologies, societal impacts, objectives, methodologies, testing frameworks, stakeholders and legal aspects, as shown in (EC CCAM Platform, July 2021). This diverse realm facilitates a wide range of research topics, in many cases attracting researchers beyond transport. The ability to find, access, exploit, use or refer to CCAM-related research data is beneficial to all interested researchers and needs to be enhanced in favor of all parties involved (EC Status of CCAM progress, April 2024). The development of CCAM systems is changing the way mobility is experienced in urban and rural areas. The technologies used rely on a continuous exchange of data between a variety of platforms, sensors and stakeholders. However, this growing complexity also brings new challenges in terms of data ownership, technical interoperability and secure information flow. Against this backdrop, the SCILAKE CCAM pilot project was launched to investigate how the principles of Open Science and FAIR (Findable, Accessible, Interoperable and Reusable) data can be implemented to support safe, inclusive and effective mobility solutions. Key regulatory instruments are shaping the direction of this change. The EU Data Act (2025) and the emerging European Mobility Data Space (European Commission, 2023) aim to ensure fair and equitable access to data while fostering innovation between public and private actors. Similarly, global standards such as the WP.29 regulations (UNECE, 2025) define cybersecurity and data recording requirements for automated vehicles, while ISO 22737 (2021) sets out specifications for lowspeed automated driving systems, highlighting the need for technical standardisation and secure communication protocols. These frameworks emphasise the regulatory push towards open, ethical and interoperable CCAM systems. Nevertheless, the practical implementation of FAIR data in CCAM remains a challenge. One of the most pressing issues is the unclear allocation of data ownership and access rights, especially between vehicle manufacturers, platform providers, authorities and third-party service providers. In addition, ensuring strong data protection while sharing data requires robust encryption, anonymisation and legal protection. Technical fragmentation makes integration even more difficult, as mobility data is generated and processed in very heterogeneous systems. The CCAM pilot project addresses these challenges by applying FAIR data practices to enable transparent, secure and scalable data exchange between CCAM stakeholders. It explores governance models and technical solutions, such as interoperable APIs, privacy-friendly algorithms and harmonised data sharing agreements, that support real-time collaboration while respecting legal and ethical boundaries. With this approach, the pilot project aims to demonstrate how the principles of Open Science can accelerate innovation in CCAM while ensuring accountability, user trust and policy alignment. Open Science in transport research: insights from the Maritime and CCAM pilot cases 6 3.3 Methodology followed The two expert communities participated in the co-design process with the technical partners on the tools to be developed, prior the pilot phase. Initially they were provided with a detailed questionnaire created by the technical partners. It concerned their engagement in existing science knowledge graphs and the use of open data types, monitoring the established knowledge and the individual research trends. Specific results on knowledge bases, scientific graphs and proper data types were provided along with details and references wherever applicable. This first loop was then acknowledged by the tools and services implementors, who in further iterations received more precise information regarding the data model, the main domain entities of interest and their interconnection. The data model was designed according to the needs of each pilot, identifying knowledge domain entities of research interest. Those entities were represented as nodes in a graph, the scientific graph, while their interconnections represented the edges. Those compatible to the OpenAIRE scientific graphs formed the basis for the creation of personalized solutions for the pilots. For example, classes such as vessel type, regulation and publication were integrated into the graph for maritime transport (figure 1), while the CCAM graph (figure 2) focused on technologies, standards and physical and digital infrastructure. Figure 1: SKG-model diagram for Maritime Transport pilot During the pilot phase, curated datasets were read, mapped and queried via LakeAPI and AvantGraph, with a focus on semantic enrichment and provenance tracking. Although no quantitative performance comparison was carried out, the iterative feedback from users (via structured tool walkthroughs) was incorporated into further improvements. As part of the pilot for the maritime sector, data on publications, regulations and vessel types were processed and more than 27,000 research products on important maritime concepts were created. Similarly, for the CCAM pilot, over 11500 research products were extracted from mobility research using NLP pipelines for sentiment classification and stakeholder relevance mapping. Preliminary qualitative feedback emphasised the improved findability of policyrelevant publications and easier navigation through cross-cutting data units (e.g. linking funding bodies to regulatory guidelines or transport datasets). Figure 1 and Figure 2 illustrate the maritime and CCAM schema structures that reflect relationships such as dataset regulation, author organisation and Vessel Open Science in transport research: insights from the Maritime and CCAM pilot cases 7 Type emissions. These were derived from the specific SKG models and mapped to OpenAIRE and EOSC compatible standards. Figure 2: SKG-model diagram for CCAM pilot 3.4 SCILAKE solutions Both the Maritime and CCAM pilots utilised SCILAKE’s modular architecture with Tier I - Core Platform Services” and Tier II - Value-added & User-Facing Services (figure 3) to support the implementation of FAIR data and Open Science practices in transport research. These services were selected not only for their technical capabilities, but also for their alignment with the core values of transparency, inclusivity and interoperability. As for SCILAKE's TIER I services, AvantGraph played a key role in both pilot projects by enabling explainable graph analyses. In the case of maritime transport, AvantGraph was used to visualise complex relationships between different data sources and helped users to validate the results of data integration processes. In the CCAM pilot project, AvantGraph enabled stakeholders to understand the relationships between mobility datasets, regulatory frameworks and system components, providing transparency in automated decision chains. AvantGraph's customisable algorithms ensured a balanced treatment of the different datasets, taking into account fairness and domain-specific requirements. Lake API served as the technical backbone for interoperable data access. By adopting the Scientific Knowledge Graph Interoperability Framework (SKG-IF), it ensured standardised access to various Open Science in transport research: insights from the Maritime and CCAM pilot cases 8 datasets in both pilots. In the maritime sector, it enabled structured queries via knowledge graphs that include operational, legal and environmental maritime data. In CCAM, it facilitated the integration of sensor data, vehicle-to-infrastructure datasets and standards so that stakeholders could retrieve and reuse relevant data according to standardised rules. Figure 3: SCILAKE services For SCILAKE's TIER II services, BIP! Spaces was used in both pilots to visualise domain-relevant knowledge and support equality in data visibility. In the maritime context, BIP! Spaces enabled the ranking and filtering of freely accessible research results related to the digitalisation of ports and the optimisation of logistics and helped to highlight underrepresented ports and smaller research contributions. In CCAM, it supported the exploration of new research trends in the areas of connected mobility, autonomous systems and regulatory innovation and ensured that newer or less cited work was not overlooked. In addition, the SciNoBo Suite improved fairness and adaptability in the classification and presentation of knowledge. In the maritime domain, it supported the analysis of citation intent and polarity of studies related to maritime safety, regulatory compliance and logistics, contributing to a balanced presentation of technical and policy research. In CCAM, SciNoBo’s provenance tracking and feedback loops helped align findings with stakeholder needs, enabling dynamic adaptation of results based on user feedback and research context. Figure 4: OpenAIRE gateway (https://beopen.openaire.eu/) An additional service is provided by the OpenAIRE gateway (figure 4), which summarises open access research results that are relevant to both the maritime and CCAM sectors as indicated by the two pilots Open Science in transport research: insights from the Maritime and CCAM pilot cases 9 accordingly. In the maritime sector, OpenAIRE gateway focuses on studies on digital twin port systems, maritime cyber security and the standardisation of AIS data. These results (enumerating 27 K relevant research products) complemented the objectives of the maritime pilot project by emphasising the importance of explainable analyses and structured data queries via Lake API and BIP! Spaces. For the CCAM pilot (enumerating 11.5 K relevant research products), the Gateway highlights ongoing efforts in the areas of data governance, cybersecurity in automated vehicles, and public trust, to which the SCILAKE tools, particularly AvantGraph and SciNoBo, have contributed by mapping regulatory interdependencies and assessing sentiment in the research discourse. The alignment between the SCILAKE architecture and the OpenAIRE gateway illustrates the value of federated, transparent research infrastructures that enable FAIR-orientated transport innovation in all domains. 3.5 Challenges in implementing FAIR data The implementation of FAIR data in the transport sector remains a complex challenge, characterised by the diversity of stakeholders, legal restrictions and technological fragmentation. One of the most persistent obstacles is the lack of effective collaboration between key stakeholders. While academic and public research organisations advocate for open access and data reuse, private sector players, such as OEMs, tier 1 suppliers and logistics companies, are often reluctant to share proprietary information, including datasets, testing results and internal methodologies. This divergence in the culture of data sharing hinders comprehensive, cross-sector knowledge integration. In both pilot cases, researchers are often faced with the additional burden of navigating dispersed, non-interoperable information systems. Finding relevant data in dispersed repositories can be time-consuming and inefficient. Even when datasets are identified, they may come with unclear licencing, limited documentation or restrictive access conditions. These challenges delay the application of FAIR principles and reduce the potential for reproducible and impactful research as presented in table 1. Table 1: FAIR Data Implementation Challenges Challenge Area Maritime Pilot CCAM Pilot Stakeholder Collaboration Isolated data among port authorities, shipping companies, and regulators Private sector (OEMs) resistant to sharing testing and vehicle data Regulatory Complexity Overlap of IMO, EMSWe, and GDPR requirements Compliance with Data Act, WP.29, and ISO 22737 Technical Fragmentation Incompatible systems across logistics and maritime operations Heterogeneous sources (vehicles, sensors, V2X systems) Data Ownership & Access Unclear rules for proprietary vs. public operational data Conflicts over data generated by public vs. private systems Information Discovery Difficult to locate standard port and vessel datasets Mobility data scattered across diverse platforms