Periodic Reporting for period 1 - CONNECTOR (CustOms exteNded iNteroperablE Common informaTiOn shaRing environment)
Période du rapport: 2023-10-01 au 2024-09-30
The goal of Integrated Border Management is to manage the Schengen external borders effectively and efficiently so as to make the Schengen area more secure. This will be achieved by better control means, while remaining as open as possible to legal travelers. However, there are interoperability issues between Customs & Border/Security IT systems that were highlighted in several relative assessments (the Assessment report of practitioners, 2019 and in The Import Control System, ICS2) and its relevance as an EU internal security system, 2021, where according to the Customs Authorities, EIS and SIS data could be of value for them. To this direction, the EC published the EU Customs Action Plan (2021-2027) to technologically and structurally enhance Customs, through financial support under the Customs Control Equipment Instrument.
Furthermore, the briefing of the Committee of Budgetary Control to the EU Parliament concluded that ‘There is not a harmonized shared risk assessment, management, and control system for all Member States. In practice, customs administrations apply their own risk assessment & management systems, without even using the same IT system. This can weaken coordination and hinder effective sharing of information. Establishing interoperability among their IT systems is of paramount importance.
CONNECTOR’s vision is to contribute to the European Integrated Border Management and to the EU Customs Action Plan by i) addressing the need of close cooperation between Customs, Border & Coast Guard Authorities within the current and upcoming challenging & demanding environment of borders’ control, and ii)further involving Customs to the Common Information Sharing Environment network and the Enhanced Common Information Sharing Environment (e-CISE), via the development of the Customs Extended Common Information Sharing Environment (CE-CISE).
In addition, CONNECTOR aims to develop, for the first time, an integrated, common and shared risk assessment approach for all IBM Authorities, considering the pan-EU common risk indicators per end user group (Customs, Border and Coast Guards Authorities including FRONTEX).
This will serve to secure external EU borders, protect EU citizens from cross-border crime and/or secure the seamless flow of travelers, as recommended in the multiannual strategic policy document. The CONNECTOR system will be developed as an interoperable technical environment, ensuring close cooperation and information exchange at all levels. The design & development of the CONNECTOR system will be based on the analysis of current policy initiatives in EU level (directives, policy documents, etc.) along with gaps, needs and future views of the end-user groups, going beyond previous initiatives (ANDROMEDA, MARISA, EFFECTOR, etc.), and complying with the Societal, Ethical and Legal requirements and regulations. The CONNECTOR system will be validated in real operational environment, based on well-defined National, Crossborder and Transnational use cases, commonly defined by Customs, Border and Coast Guards authorities, during three long lasting trials (Demonstration & Testing) under a standardised methodology.
Furthermore, the briefing of the Committee of Budgetary Control to the EU Parliament concluded that ‘There is not a harmonized shared risk assessment, management, and control system for all Member States. In practice, customs administrations apply their own risk assessment & management systems, without even using the same IT system. This can weaken coordination and hinder effective sharing of information. Establishing interoperability among their IT systems is of paramount importance.
CONNECTOR’s vision is to contribute to the European Integrated Border Management and to the EU Customs Action Plan by i) addressing the need of close cooperation between Customs, Border & Coast Guard Authorities within the current and upcoming challenging & demanding environment of borders’ control, and ii)further involving Customs to the Common Information Sharing Environment network and the Enhanced Common Information Sharing Environment (e-CISE), via the development of the Customs Extended Common Information Sharing Environment (CE-CISE).
In addition, CONNECTOR aims to develop, for the first time, an integrated, common and shared risk assessment approach for all IBM Authorities, considering the pan-EU common risk indicators per end user group (Customs, Border and Coast Guards Authorities including FRONTEX).
This will serve to secure external EU borders, protect EU citizens from cross-border crime and/or secure the seamless flow of travelers, as recommended in the multiannual strategic policy document. The CONNECTOR system will be developed as an interoperable technical environment, ensuring close cooperation and information exchange at all levels. The design & development of the CONNECTOR system will be based on the analysis of current policy initiatives in EU level (directives, policy documents, etc.) along with gaps, needs and future views of the end-user groups, going beyond previous initiatives (ANDROMEDA, MARISA, EFFECTOR, etc.), and complying with the Societal, Ethical and Legal requirements and regulations. The CONNECTOR system will be validated in real operational environment, based on well-defined National, Crossborder and Transnational use cases, commonly defined by Customs, Border and Coast Guards authorities, during three long lasting trials (Demonstration & Testing) under a standardised methodology.
The CONNECTOR project has accomplished the following objectives :
On the operational landscape analysis: 1)Analysis of the existing landscape starting from the consortium end users, identifying systems, data, assets of interest for common operations 2) Screening of the legal & ethical frameworks to identify technological barriers to lead the technological approach 3)Identification, gathering and prioritization of existing needs, gaps & operational requirements for Customs and Border/Coast Guards in the EIBM domain 4)Analysis of the differences and commonalities of RA and DSS tools, developing an approach for a common situational picture 5)Delivering of high level scenarios and assets, resources, data, etc. to lead next WPs, prior to the detailed descriptions the second Reporting period.
On the Technical design and High Level Architecture:
1)Definition of the functionalities and forming of the technical specifications of the system 2)Design all the necessary protocols and hardware for the interconnection of the adaptors 3)Design of a complete and effective early warning system that supports three main functions: risk analysis, monitoring and decision support 4)Design of data fusion models and algorithms and building of the Digital Twin prediction models that will forecast illegal activities 5)Design of the system architecture
On the operational landscape analysis: 1)Analysis of the existing landscape starting from the consortium end users, identifying systems, data, assets of interest for common operations 2) Screening of the legal & ethical frameworks to identify technological barriers to lead the technological approach 3)Identification, gathering and prioritization of existing needs, gaps & operational requirements for Customs and Border/Coast Guards in the EIBM domain 4)Analysis of the differences and commonalities of RA and DSS tools, developing an approach for a common situational picture 5)Delivering of high level scenarios and assets, resources, data, etc. to lead next WPs, prior to the detailed descriptions the second Reporting period.
On the Technical design and High Level Architecture:
1)Definition of the functionalities and forming of the technical specifications of the system 2)Design all the necessary protocols and hardware for the interconnection of the adaptors 3)Design of a complete and effective early warning system that supports three main functions: risk analysis, monitoring and decision support 4)Design of data fusion models and algorithms and building of the Digital Twin prediction models that will forecast illegal activities 5)Design of the system architecture
Data Fusion: Multimodal fusion with Graph Neural Networks is a powerful approach to integrating diverse information sources, that can enhance maritime traffic safety. It can for example help detect abnormal ship behaviour, identifying dangerous manoeuvres, unauthorised activities, and illicit practices. Advancements in GNN architecture, attention mechanisms, and dynamic graph handling are pushing the boundaries of multimodal data, unlocking new possibilities and driving innovation across various applications. The Data Fusion Tool will apply these solutions in order to enhance the assessment of the situations that the CONNECTOR platform will encounter.
Additionally, the Predictive Analysis (PA) algorithms that are being developed within the digital twin framework focus on forecasting illegal activities occurring in customs and at sea. While current methodologies in this field have significantly advanced through the application of machine learning, deep learning, and data fusion techniques, our implementation enhances these methodologies by integrating abnormal ship route detection techniques with additional information from a diverse range of data sources, including crew and customs data, expanding existing predictive models.
Knowledge Graph and NLP algorithms: Enhanced semantic enrichment through advanced models allows for better disambiguation and context representation, especially in domain-specific texts. Dynamic ontologies are integrated to adapt to evolving data, while advanced graph-based representations capture multi-dimensional relationships enabling a common risk management approach. Scalable data integration and NLP-based document analysis enables extracting valuable information and linking diverse sources. The tool supports reasoning providing input to the DSS for actionable insights for decision support. Finally, integration with Risk Data Sharing Architecture facilitates automated, cross-border information sharing for high-risk activities. These advancements result in a dynamic, semantically enriched knowledge graph that improves decision-making and analysis in sensitive domains like border risk management.
Additionally, the Predictive Analysis (PA) algorithms that are being developed within the digital twin framework focus on forecasting illegal activities occurring in customs and at sea. While current methodologies in this field have significantly advanced through the application of machine learning, deep learning, and data fusion techniques, our implementation enhances these methodologies by integrating abnormal ship route detection techniques with additional information from a diverse range of data sources, including crew and customs data, expanding existing predictive models.
Knowledge Graph and NLP algorithms: Enhanced semantic enrichment through advanced models allows for better disambiguation and context representation, especially in domain-specific texts. Dynamic ontologies are integrated to adapt to evolving data, while advanced graph-based representations capture multi-dimensional relationships enabling a common risk management approach. Scalable data integration and NLP-based document analysis enables extracting valuable information and linking diverse sources. The tool supports reasoning providing input to the DSS for actionable insights for decision support. Finally, integration with Risk Data Sharing Architecture facilitates automated, cross-border information sharing for high-risk activities. These advancements result in a dynamic, semantically enriched knowledge graph that improves decision-making and analysis in sensitive domains like border risk management.