The MELCHIOR project is dedicated to the development and demonstration of a very specific technology: infrasound interrogation called Z-MESMERISE.
The project’s homepage says that this is “a novel technology for fast detection of drugs, explosives, weapons and illicit goods concealed on individuals and in critical cavities of the human body based on infrasound mechanical impedance interrogation, optionally complemented with other harmless and non-contact technologies.”
MELCHIOR promises “to drastically increase the safety of the border staff having to touch inspected individuals. Thus, travelers’ experience will improve greatly.”
As a consequence of the fact that “no contact is needed” the conclusion is drawn that “individuals do not have to undress and therefore their privacy is respected.”
Specific objectives include “to design and implement an AI able to detect concealed items under clothes based on the information collected by the sensors,” and “to train the AI algorithms to identify items and perform correct discrimination of threats.”
A video presentation is also available.
MELCHIOR’s technological outputs are described (no technical deliverable was publicly available at the time of writing, ed.) in the project’s About section, as follows:
“The present Project aims to develop up to TRL7-8 blast-proof prototypes with detection capability extended to limbs and body cavities and improved sensitivity. Infrasound computer aided modelling, improved Artificial Intelligence, altogether with improved new infrasound generators and sensors will enable an improvement in reliability and sensitivity. Combination with other non-contact technologies like low-cost millimeter waves and THz (terahertz, ed.) point echo sensors with no image formation will be explored. Additionally, compatibility and complementary combination of specific bioaerosols and chemical vapour detection novel technologies will be explored in the prototypes as well.”
According to the second project Newsletter from April 2024, “this combination of technologies being developed in the MELCHIOR project has the potential to deliver significant advancements in security and operator safety at airports, large stadia and in prisons right across Europe and beyond.”
MELCHIOR is one of the BES (Border External Security) Cluster projects headed by METICOS.
Also, its technological outputs build on those already obtained by previous EU-funded projects — at an inferior technological readiness level. For example, “Mechanical Impedance interrogation was validated in lab -TRL4- in the multiply awarded H2020 MESMERISE Project, detecting items concealed on trunk under clothes -including critical cases like warm molded materials that remain undetected to other technologies.”
The very idea for the MELCHIOR project came up in the MESMERISE project because a team from the University of Alcalà “had been working with infrasound, echolocation, and other uses of sound.”
There were “synergies” with several EU-funded projects, including NOTIONES (“The H2020 NOTIONES project aims at establishing a network of academia, industry, security and intelligence practitioners to monitor technological opportunities and contribute to the definition of further security and intelligence research”), BORDERUAS, NESTOR, METICOS, ODYSSEUS, FLEXI-cross, and others. As common in EU-funded projects involving cross-border security, everything is about efficiency: “Together, we’re pioneering cutting-edge solutions to safeguard borders with unprecedented efficiency and effectiveness,” was published on ODYSSEUS’ X account when announcing “new synergy” with MELCHIOR.
Other synergies include METEOR (“The METEOR project’s goal is to revolutionize cargo screening with innovative and efficient technology capable of delivering fast, reliable, and economical container inspections”) and TENACITY. TENACITY is about “an interoperable open architecture for the integration and analysis of multiple data from a variety of sources, by exploiting game changing digital technologies.”
While we lack substantial details concerning project pilots, we do know that “slight variations of the prototypes will be demonstrated in operational environments (TRL7) of customs and border authorities in EU external air, land and sea borders crossing points in Spain, Greece, Romania and Finland. Demonstrations in other operational environments like an arena or prisons will increase and broaden the impact on security of society.”
Five tests “will take place in the final six months, between March and August 2025. These Tests will provide truly tangible opportunities for our customs and border control partners to use and assess the MELCHIOR technologies. The technologies that have been developed and the final version of the prototype blast-proof cabins will be demonstrated in operational environments of our partner customs and border authorities in European external air, land and sea borders crossing points in Spain, Greece and Romania. Because the multiple benefits and enhanced security offered by the technical solutions extend into other operational environments, for example arenas and prisons, demonstrations will also be held in these locations in Spain and Finland.”
No project deliverable was publicly available after two years project runtime — both on MELCHIOR’s website and on its Cordis page. This means that there were no technical details accessible.
Consequently, all major concerns associated with the use of AI in such contexts apply: fairness, privacy, human rights, and structural opacity.
MELCHIOR also excessively focuses on technological solutions to optimize the management of cross-border flows, and the efficiency of procedures, which Newsletter #2 exemplifies: “Perhaps the biggest benefit for customs officers will be MELCHIOR’s development of a new tool that will not disrupt the flow of people. In fact, it will increase that flow with a relatively high rate of reliability. If it normally takes one minute to physically search a person and you can do this in half that time without touching the person by using our technology, you are making everything smoother, less intrusive, simpler and more efficient, with the added level of respect for the physical privacy and dignity of the person being searched.” Is a “relatively high rate of reliability” an acceptable threshold — both technologically and morally? Can mistakes be justified with time and cost savings?
The very newsletter touches on such concerns: “For sure we have to be aware of and vigilant about any potential misuse of Artifical Intelligence. As far as AI is concerned, like all science, it must comply with ethics. (Project coordinator, ed.) Dr. José Luis Pérez Díaz concluded that ‘if you forget humanities, you will become non human. That is a real problem’.”
