Border surveillance is a complex mission given the length of European borders and the variety of terrain configuration. Historically, the Mediterranean region has always been witnessing challenges like migration flows, cross-border crime, persistent natural hazards (forest fires, floods and earthquakes, etc.), which required cross-border and multi-discipline collaboration in order to optimise response actions. Several efforts have been made in the past and current years to increase the efficiency of surveillance activities via the collection and exchange of land and maritime surveillance information between control authorities – even across national borders. In 2010, the European Commission put forward a six-step Roadmap towards the Maritime Common Information Sharing Environment (CISE). The goal was to create a political, organizational and legal environment to enable information sharing across the seven relevant sectors/user communities (transport, environmental protection, fisheries control, border control, general law enforcement, customs and defense) based on existing and also on future surveillance systems, networks with a view to achieve a fully operational CISE by 2021[1]. ANDROMEDA unlocked the full capabilities of the CISE by enhancing the Maritime CISE Model, extending its scope to the Land Surveillance Information Exchange and providing and demonstrating 100% compatible Command & Control, Data Fusion and Decision Support systems. The enhanced CISE (e-CISE) Model will streamline the integration with current and future operational systems and is perfectly aligned with the overall European policy that facilitates the interagency interoperability and cooperation and allows each Member State to decide how, when and whether additional data sources are of relevance to its operations.
Data Fusion (DF) services are cornerstone elements in maritime and land border surveillance systems. The Data Fusion services of ANDROMEDA aim to enhance the situational awareness of an organization by producing higher level and more accurate information, by processing several data sources and uncovering hidden patterns, correlations and other useful intelligence. The whole Data Fusion workflow of ANDROMEDA is able to combine both maritime and land domain information, thus providing the end-user with a complete operational picture about the field of operation. In terms of interoperability, the DF services are interconnected each other via a dedicated messaging middleware provided by EXUS, which is the DF Gateway. Regarding the information exchanges between the DF services, the adopted approach is fully compliant with the eCISE data model which allows ANDROMEDA to easily consume information from other CISE-based networks.
The Data Fusion Gateway (figure 1) facilitates the messaging infrastructure for exchanging eCISE information between the different DF services and also offers storage and monitoring capabilities. The design of the DF Gateway is based on the microservice architecture which has several advantages over monolithic applications. However, the main reason that we followed this approach in ANDROMEDA is the benefits of scalability. Having in mind that ANDROMEDA may be extended in the future, either by adding new functionalities and services or data sources to be consumed by the existing services, we offered a scalable messaging framework.
[1] Official Presentation of the Andromeda Horizon 2020 project (https://www.andromeda-project.eu/downloads/index.html)
Figure 1: Data Fusion gateway
Towards this direction, EXUS also developed a messaging infrastructure offering the main functionality of the DF Gateway and aiming to provide the DF services with an efficient way to exchange eCISE messages in real-time. Having also in mind that we should ensure that the deployed message bus should offer some additional benefits besides real-time messaging, such as high availability and scalability in terms of performance. This is achieved by deploying an Apache Kafka[2] cluster to act as the communication middleware among the DF services. On top of the above the DF Gateway provides monitoring capabilities
[2] Apache Kafka (https://kafka.apache.org/)
[3]Prometheus (https://prometheus.io/)
[4] Grafana (https://grafana.com/)
Figure 2: DF Gateway graphs using Grafana
EXUS offered also the Early Warning Engine (EWE), a Data Fusion service responsible for the threat assessment of ANDROMEDA, that produce meaningful alerts which are critical to the operator. It leverages the existing tracks (e.g. land vehicles, vessels) generated by the lower-level DF services and combines them to construct the state of the operational field. EWE continuously monitors tracks which are subjects to previously generated anomalies and identifies critical patterns, enhancing the intelligence provided to the operator of the C2 systems. The patterns identified by EWE are transformed to alerts in terms of e-CISE entities and are transmitted to the ANDROMEDA C2 platforms via the Data Fusion Gateway messaging infrastructure.
Figure 3: EWE visualization module showing anomaly clusters
The main advancement of the Early Warning Engine was to extend the existing functionality to also support JDL[5] 3 Data Fusion regarding Threat Assessment and provide a tool which incorporates data from all possible sensors and surveillance systems in various fusion levels that ANDROMEDA is currently targeting. This was achieved by enforcing ML algorithms for extracting intelligence from anomalies that were generated by the lower-level DF services and producing meaningful alerts. Also, the Early Warning Engine has been extended to be fully-compliant with the e-CISE data model and to support both land and maritime information. The software implementation is based on the microservice architecture, thus allowing us to be scalable and easily deployable in modern deployment environments using Docker containers.
This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 833881. This article reflects only the authors’ views and the Research Executive Agency (REA) is not responsible for any use that may be made of the information it contains.
Comments