SEAGUARD aims to enhance “real-time maritime situational awareness technologies” through “an innovative and integrated maritime cross-border surveillance system” which “will combine multiple sensing technologies across fixed and mobile platforms”. It will feature “unmanned vehicles (UxVs), fixed buoys, submarine cables [described as “(SMART) submarine cables”], and advanced analytics to monitor and detect various threats, such as illegal border crossings, contraband smuggling, illegal fishing, and terrorist activities”.
The system promises full interoperability with border surveillance capabilities from all member states, “ranging from the Mediterranean to the Nordic Countries”, while also being “scalable, flexible and relocatable, since the needs along borders shift rapidly in time, with changing and improving threats”. In fact, flexibility is a key requirement of the Horizon Europe call that funded the project, according to which “More efficient and more flexible solutions (including for relocation, reconfiguration and rapid deployment capabilities) than physical barriers to deter and monitor irregular border crossings outside border crossing points” is one of the “expected outcomes”.
Its official website summarizes SEAGUARD with an ambitious claim: “Unmanned AI Robotic technology changes the future of maritime border control”.
SEAGUARD promises “advanced solutions” to integrate “different types of sensor technologies”. These will amount to “unmanned vehicles (air, surface and underwater) and intelligent underwater cables, combined with advanced analytics and decision-support systems”. The project’s website cryptically adds that all the information collected through these tools “will be merged and processed by classification algorithms, and then made available to end users, who will be able to make educated decisions and take action”.
We tried to better understand the envisioned system and how it works, by asking the EU Commission’s Research Executive Agency for SEAGUARD’s “Grant Agreement”. We were however (unsurprisingly) met with extreme redactions, which remained even after we challenged an initial release that basically amounted to a fully redacted document.
We did however learn that SEAGUARD is envisioned as “a system of federated swarming systems”, i.e. — according to one of the only readable passages in the document we obtained — a system of systems “where individual sub-swarming C2s (“Command and Control” systems, i.e. computer solutions “to manage a single surveillance system” or “to oversee multiple systems that are dispersed across a large area”, ndr) are under the full control of an upper level C2 responsible for the cross-border surveillance mission, whilst every vehicle in mission is considered as its sensor with specific tactical capabilities that vary dynamically“. Tough to say what this military-inspired jargon actually translates to, but the basic idea seems to be that of having independent autonomous sub-systems that can be all (autonomously?) coordinated by a higher-level authority, in a variety of settings.
SEAGUARD directly builds on previous EU-funded projects I-SEAMORE (an integrated maritime surveillance system based on unmanned vehicles and AI) and EURMARS (a new vision for a multitasking, AI and UxVs-based maritime surveillance platform), by design. This is confirmed by a EU Commission presentation given in October 2024, containing a slide that also illustrates further relationships with projects such as the recent BORDERFORCE, and with past endeavours such as ANDROMEDA, COMPASS2020, EFFECTOR and NESTOR.
Currently unknown. We do know, however, that “The project encompasses the design, development, testing, and deployment of a comprehensive surveillance system that integrates multiple sensing technologies on both fixed and mobile platforms”.
More concretely, a research paper “partially funded” under SEAGUARD demonstrated “how to implement a cooperative multi-modal robotic system for coverage path planning involving a UGV (unmanned ground vehicle) and a UAV (unmanned aerial vehicle), both “in simulation and in real life”. As a result, “The robots can cooperate to produce effective exploration, and a set of plans was developed using a robust coverage path planning technique”, wrote the researchers. In the paper, “coverage path planning” is defined as “the process of determining an optimal path for a robot to reach every point in an environment”. While “enabling continuous exploration” is a key objective, environmental sustainability is not mentioned.
Besides the usual solutionist premise according to which “Enhanced border surveillance across EU borders is essential for monitoring and deterring illegal activities, particularly through real-time maritime situational awareness technologies”, it is currently difficult to say what this project actually amounts to.
In its initial reply, REA vaguely defined SEAGUARD as involving “technological solutions aimed at countering illegal activities, which may include maritime security, border protection, and law enforcement operations”. It contextually shared an almost entirely redacted “Grant Agreement” in which not even headlines, working packages or deliverables were readable. We could only learn that some are to be considered “EU RESTRICTED”, and therefore will only be handled by the EC, Frontex and Consortium Members. Milestones and critical risks were also crucially redacted.
We tried to challenge the REA’s decision through a “confirmatory application”, but the Agency’s director, Marc Tachelet, confirmed that this level of censorship was necessary to protect public security, privacy and the commercial interests of Consortium members.
There is no public interest to be safeguarded here, argued Tachelet, and no transparency is due neither on the envisioned use cases nor concerning the methodologies applied over the course of the project. Wrote Tachelet in his reply to our “confirmatory application”:
“The use cases, methodology, and operational contexts contain information that could expose vulnerabilities in security infrastructures and detection mechanisms or reveal strategic capabilities that could be exploited by adversaries or criminal entities. It could also undermine the effectiveness of SEAGUARD by allowing hostile actors to adapt and circumvent the proposed security measures and compromise classified or sensitive collaborations with governmental and defence agencies. The exposure of this information could pose a direct threat to the integrity and effectiveness of security and defence operations.”
This, concluded the REA director, equally applies to “the list of work packages, deliverables, milestones and critical risks”. Even knowledge of a mere title of an official document to be produced over the course of SEAGUARD (and any other project, really) can and should be prevented in order to safeguard the participants’ intellectual rights and public security. It is obvious that this criterion can be arbitrarily applied to any transparency requests, effectively rendering the whole process moot.
The amount of project participants with clear ties to the defence sector and the military is also concerning, and could help explain the extraordinary amount of opacity that surrounds the project even after we exhausted the legal access to information request procedure. Besides Portugal’s Defence Ministry, which is part of the Consortium, members include multiple private entities at work on unmanned solutions for the Defense sector. These include Whitesteps Technologies, Utek (“covering naval, ground and aerial solutions for both defense and civil applications”), Magellium (which provides “its customers with images processing and technical, software and scientific engineering solutions, for civil or military uses”) and ARQUIMEA. AZUR DRONES claims to be “the European leader in drone surveillance and inspection solutions”, and boasts: “As a pioneer of drone-in-a-box infrastructure, we design a fully automated solution capable of operating in the most sensitive and complex environments”.
The Consortium also includes UNMANNED TEKNOLOGIES APPLICATIONS SL (SPA), or UTEK, a USV provider (motto: “real unmanned systems for real needs”, claims a video presentation) that boasts to have “performed several demonstrations and participated in different projects and military exercises acquiring a solid operational background. Missions were executed under the supervision of various civilian and military entities.” According to a 2020 video presentation (cfr. above) it operates, among others, in the “defense” sector. The video (0:36) shows a USV equipped with a weapon on a turret. The video actually goes further than that, showing “defense solutions” that include
“search, sweeping and neutralization” of mines, “anti-submarine warfare”, “surface warfare”, “electronic warfare” and “armed scort”.
All these companies received EU funding to develop “exclusively civilian” applications.
