Andromeda aims to realize a better and extended information sharing environment for both maritime and land border surveillance.
In particular, the project intends to build on both EUROSUR (“the European Border Surveillance System” that “has the objective of improving the situational awareness and increase reaction capability at external borders the Union’s especially its southern maritime and eastern land borders”) and CISE (the Maritime Common Information Sharing Environment born in 2010 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”).
The aim is “to unlock 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.”
Exploitable results include:
1) Command and Control Systems, such as the ENGAGE BME platform “for first responders in the field and personnel in a local or regional or national command centers”, or the Socrates C2 solution “for local, regional or national command centers”;
2) Data Fusion systems, such as the TRITON Abnormal Vessel Behaviour Engine, “an abnormal vessel behaviour service that uses a geospatial complex event processing to identify and analyse motion patterns which was enhanced with additional monitoring patterns for maritime and includes also the detection of additional anomalies in the land domain”, or the ANDROMEDA Data Fusion Engine — “a platform which performs Data Fusion operations on data, originating from different sources and combines them to produce accurate and meaningful high-level information”;
3) Decision Support Tools, such as OCEAN-SAR, a service “to support maritime authorities and operational centers during search-and-rescue operations” and IANUS, a specialization of the ESSG – Enterprise Service Security Grid – Framework to provide “enhanced classification of targets based on their behaviour, behaviour analysis, comparison with learned models, identification of common and divergent behaviours, automatic generation of alerts associated with targets that present underlying threats or offenses”;
4) Advanced Assets such as the INUS Platform, an “Intelligent UxV Surveillance platform for object detection, identification and tracking via UAV and terrestrial image processing providing e-CISE interface for data sharing with C2s in ANDROMEDA.”
“The project will leverage on the developments, results and experience of the consortium from current and previous research projects (PERSEUS, CloseEye, MARISA, RANGER)”, writes ANDROMEDA’s Cordis page.
Deliverable D2.4 adds that “In the current environment where governments in the EU face existing and continuing budget pressures, the need for cost effective solutions is of paramount importance.” Consequently, ANDROMEDA “aims to make use of the capabilities and results of other European programs (…),” as “the main EU Industry and Practitioners involved in the central Security European programs (e.g. EUCISE2020, Perseus, CLOSEYE, MARISA, RANGER, EWISA, CoopP etc.) are present in the consortium composition.”
Deliverable D1.4 lists SMILE, D4FLY, MEDEA, RANGER and MARISA in the project’s “User Community”.
ANDROMEDA has also been deemed “of interest” by Frontex in 2020.
Three trials are described in the project’s final report (D1.4):
1) The “Iberian Maritime Border Trial” was “performed on the 7th May 2021. The area monitored for the trial was the South Coast of Portugal (Algarve), and the Italian Waters on the Mediterranean. The trial was an inter-country, inter-sector Maritime Border trial involving two EU agencies (the Italian Navy and the Portuguese Navy), with two scenarios demonstrating joint activities on Maritime Traffic Control, Illegal Activities, Smuggling and Drugs”. However, “Due to Covid-19 limitations, the trial was performed remotely and with mostly simulated data”;
2) Greece-Bulgaria Land Border – Maritime Trial: “The task involved the deploying of systems to end-user premises (Hellenic Coast Guard, EAMA, Hellenic Police, Hellenic Navy), training and supporting the end-users during their usage of the system for four releases beginning from July 2020 and ending June 2021”. Importantly, the deliverable specifies that “several assets and sensors have been deployed in the borders that will remain functional and operational after project closure”. These include X-band surface IP radars and long-range IR and thermal cameras for — among others— an 18m mast, a river patrol boat and fence poles in the Evros Delta region. A “mobile Command and Control center” was also “deployed in the land borders with a long-Range IR – Thermal camera in order to cover the surveillance of a totally remote and inaccessible area.”;
3) Ionian-Adriatic Maritime Border Trial, dealing with “illegal immigration, search & rescue, illegal fishing, and smuggling in maritime areas”, with the aim of proving data sharing — alerts, anomalies, targets, and more — in a “long-time trial session”.
In a thread on the X platform, independent border surveillance researcher Lena Karamanidou documented how technology developed within ANDROMEDA, and first tested during the Greek trial on the Delta of the Evros river, was then quietly deployed there. Wrote Karamanidou that “the pylon and its equipment, the patrol speedboat as well as platforms developed by ANDROMEDA continued to be used by the local police after the project ended.”
Karamanidou confirmed this by analysing footage aired on Greek TV in which the head of a border management centre involved in ANDROMEDA unknowingly revealed details on what surveillance tools are actually in place in the region now — and they perfectly match with those from the trial. “Links between EU-funded research and border surveillance technology are well established,” concludes the researcher, “but this was the first time that I found a concrete connection between a specific project and specific tech used on the ground” (cfr. our investigation with Magazin Royale).
Ethics and societal issues are extensively tackled in deliverables such as “D.2.4 Legal, Ethical and Societal Aspects;” “D.1.5 Legal, Societal, Ethical Initial Report;” and “D.1.6 Legal, Societal, Ethical Final Report”. And yet, some issues remain.
Deliverable D2.4 offers an interesting overall justification for border surveillance projects: “Increased control and security measures are justified with the need to protect Europe against cross‐border crime” (even though “Protecting the European seas and borders should be aimed at both creating a secure maritime and land border environment, but also protecting the lives and physical and moral integrity of those at the sea and on land borders.”).
A very strong criterion is also stated: “the fundamental rights are necessary requirements that set limits to what is and what is not acceptable in EC funded security research initiatives (CISE 2013).”
And then, an even stronger one follows: “In the context of maritime and land border surveillance activities, it is important to perceive that EU fundamental rights and/or Human Rights concern not only Europeans, but all the people, including those attempting to reach Europe.”
According to D2.4, ANDROMEDA “actively promotes” fundamental rights. However, later in the document, concerning non-refoulement (a principle that is enshrined in international human rights law and that “prohibits States from transferring or removing individuals from their jurisdiction or effective control when there are substantial grounds for believing that the person would be at risk of irreparable harm upon return, including persecution, torture, illtreatment or other serious human rights violations”), the deliverable writes: “ANDROMEDA services enable tracking vessels not only on their own sea territories, but also in the high seas and the territorial waters of third countries. It is therefore technically possible that ANDROMEDA will be used to organise border control outside countries’ own borders and to redirect intercepted migrants to the coasts of third states”.
How can this misuse be prevented? It’s the main challenge: “The key challenge for the development of ANDROMEDA is thus ensuring that the rights of the already vulnerable refugees and other migrants are not further compromised for the interests of the more well-off European citizens.”
And again, it can’t seem to be decisively met: “The misuse of the whole ANDROMEDA solution is strongly linked to the business/adoption models of the ANDROMEDA, and especially as stand-alone solution. The key question is that how we can make it sure that the ANDROMEDA solution sold will be used only for the purposes it is mentioned. This has not so much to do with the technical features of ANDROMEDA and their development during the ANDROMEDA project, but rather to the business and governance modelling to be applied after the project.” One might be left wondering whether it is indeed a good, ethical idea to develop something you already know you won’t be able to control and that at the same time implies grave human rights concerns for vulnerable individuals and groups in case of misuse.
Also, the deliverable clearly states that “ANDROMEDA must substantiate that it is necessary in democratic society, and its use must be proportional to the justified goals.” This is, again, a very high bar for justification.
The project is also very careful to claim that “Operational decisions will never be made by a computer, even the most efficient one: it will always be a human who makes the final decisions. ANDROMEDA is meant to assist decision making”. But uncertainty runs through the document; e.g.: “In general ANDROMEDA is not interested in persons (in land boarders there might be interest for persons as well?)”.
WP8 concerns the ethics appraisal scheme adopted for the project, and D1.4 details that additional information was requested during the ethics review phase, in order to satisfy ethics checks.
Deliverable D1.6 concludes: “All in all, from ethical point-of-view ANDROMEDA was not specifically demanding, since, for example, none of the activities included any intrusive research, participation of vulnerable groups, or such. Nevertheless, ethics was taken seriously and neither research integrity, nor fulfilling the commitments did not pose an issue during the project. The challenges of ANDROMEDA solutions are in the future, i.e. in the exploitation of them. The ultimate threshold is related to fundamental principles on ethical use of surveillance technology. For this, ANDROMEDA has well prepared the basis, for example, with the ANDROMEDA Code of Conduct that disclose ethical principles on the use of ANDROMEDA and with the ethical requirements presented, for example, in the final chapter of this deliverable.”
