Marshall Islands Marine Resources Authority's offshore fisheries advisor Francisco Blaha at the fisheries authority's Majuro headquarters

Opinion Article | Fishing monitoring platforms and 'desktop experts': a combination that isn't always effective against IUU fishing

ARGENTINA
Friday, January 16, 2026, 00:10 (GMT + 9)

Technology applied to the monitoring and control of fishing activities at sea based on Automatic Identification System (AIS) data constitutes a fundamental tool that, if used well and its information correctly analyzed, makes a difference in terms of surveillance. But beware! Not everything is as these platforms seem to show us, nor as analysts interpret or conclude from that information. For this reason, it is important to know some of the most recurring limitations and inaccuracies among them.

By Lic. Sergio Almada

If one starts from an erroneous initial scenario or incorrect causes regarding a problem, it will be unlikely to arrive at an efficient solution. This applies perfectly to criminal activity at sea and constitutes a very important principle to keep in mind when a researcher is at a desk and behind a computer screen, looking at and analyzing the information provided by these monitoring platforms. Knowing how to interpret what they show us, analyzing and understanding the quality and relevance of the information, and seeking a contrast with reality is something that requires expertise and the capacity for validation and contextualization. Something that is frequently scarce, undervalued, or of little importance, especially when what is shown responds more to interests than to the search for truth with the necessary scientific rigor in an investigation.

Desktop AIS control system. Photo: Wikipedia

Paraphrasing Francisco Blaha, a specialist in fishing issues and an expert in validating hypotheses and providing operational filters to these remote desktop analyses: “centering the solution to problems surrounding fishing activity at sea on a single tool, such as AIS, carries the risk of excessively raising expectations about what this technology can offer.” Blaha states that the more the hypotheses and indicators used by these platforms are adjusted, and the better the on-the-ground verification of the methodology used, the more reliable the results will be and the better the tool to be used, which will provide more realistic figures on the problems.

It is worth exploring Francisco Blaha’s contributions to the conservation and sustainable management of fisheries, to monitoring and control, to decent work on board fishing vessels, and to the practical-operational vision of the activity—that which makes his perspective special and different and, for me in particular, more attractive and interesting.

Francisco Blaha presenting information to the media

Fishing activity monitoring platforms not only show current and historical information regarding the identification and dynamics of vessels but also report on the alleged activities they carry out (events), such as fishing, transshipments, encounters, navigation, anchoring, entry or exit from ports, AIS shutdown, and fishing effort, among others. For this, they use a methodology, indicators, training data, and algorithms that allow them to define, in many cases with the help of artificial intelligence, such activities. That is to say, their contribution to control and investigation is very important, but it is not for that reason determining for the understanding of everything that actually happens around the activity, nor for addressing without margin of error forms of maritime crime such as illegal fishing. It is not determining because the information they show must not only be validated on the ground but also contextualized at the local level, and for that, real knowledge of the activity and consultation with its main actors are needed.

PNA Coast Guard System. Monitoring, Surveillance and Control System of Activities at Sea of the Argentine Naval Prefecture. It has the capacity to detect NON-COLLABORATIVE vessels (which turn off their AIS) through the use of synthetic aperture radar satellite images provided by the Argentine Space Activities Agency (CONAE).

It is in this context where the figure of what some call “desktop experts” appears and, with them, that platform-expert binomial, which can be very good, but can also be disastrous. The result of the platform-expert conjunction will depend largely on the latter's ability to interpret and contextualize the information obtained from the platform, on the knowledge of the sources and algorithms it uses, on the knowledge and experience regarding the investigated activity, and on the possibility of interacting with the main actors who can validate what has been interpreted from primary sources.

UNOFFICIAL limits: differences between the official limit of the Argentine EEZ and the one used by Global Fishing Watch until a few months ago. It corresponds to the area known as the Blue Hole, where due to its richness, numerous distant-water fishing fleets converge. Click on the image to enlarge it.

Errors usually detected in these “desktop” analyses

• Few researchers validate their assumptions with the hydrometeorological data corresponding to the event they are analyzing. Blaha points out that anyone who has spent time on a ship knows that “weather is king at sea” and that what the ship can do will depend on sea conditions. It is not always the case that a ship maintaining almost the same position for a couple of days is committing an illegal activity, as it is very possible that by overlaying hydrometeorological data onto the AIS, it can be verified that it is actually weathering 50-knot winds and huge waves, where rather than an illegal activity, it is only trying to survive or ensure the safety of the vessel and its crew.

• Directly associating the port entry of a fishing vessel or a refrigerated fish carrier as an arrival to unload catches, without considering that it may also do so to receive any other type of services and failing to request information from port authorities to confirm or discard that presumption. For example, there are fleets like the Chinese one, operating at Mile 201, that transship their products on the high seas and practically never unload them in port.

• Considering the encounter of a fishing vessel with another or with a cargo ship as a transshipment of catches, when there are many other operational reasons to have encounters at sea, such as the provision of food, bait, fuel, spare parts, fishing gear, crew changes, and everything else we can think of. Analyzing the duration of the encounter can be a great help in identifying the reason.

• Assuming the limits that monitoring platforms present for the maritime spaces of coastal States without verifying that these correspond to official ones leads to serious errors. Comparative experience for several Exclusive Economic Zones in South America shows that the limits used by some of these platforms present divergences from the official ones, which often makes what actually occurs outside jurisdictional waters look like illegal fishing maneuvers. This happened for years with the limits used by Global Fishing Watch for the Argentine EEZ, extending it in the Blue Hole area up to 4.4 nautical miles into the high seas. This divergence also exists in the limit of the Uruguayan EEZ on its side with Brazil, without correction to date.

• Unequivocally associating the entry of a vessel into an area in which it is not authorized to operate with illegal fishing, without analyzing whether its dynamics and the pattern of its track are compatible with fishing activities for that type of vessel, thus failing to recognize the international right of free navigation that assists them.

• Another problem detected in studies that count port arrivals using AIS lies in the fact that poorly trained algorithms may consider movements between anchoring areas, from these to port and vice versa, as if they were new arrivals, exponentially increasing their quantity. This problem has been observed in studies on fish carrier arrivals at the port of Montevideo.

• Analyzing apparent fishing maneuvers from AIS positions without taking into account the type of vessel also leads to errors. Depending on the fishing gear a vessel uses, the speed, fishing pattern, depth, and characteristics of the operations area will be different, and even, in some cases, the time of day when the activity occurs to associate it with an activity compatible with fishing for that type of vessel.

• Directly associating the turning off of AIS with illegal activity without considering that not all fishing vessels are obliged to carry that device or keep it turned on by their flag States—to whom international regulations delegate this requirement—or that there are commercial competition reasons for doing so. Added to this is the failure to take into account that many coastal States, such as Argentina through the Coast Guard System, and various fishing activity control platforms have the capacity to detect them through satellite imagery, independently of their AIS emission.

• Finally, some machine learning algorithms on these platforms erroneously infer activities as fishing that, for an expert, do not correspond to that type of maneuver for a certain fishing gear. Such errors contribute to magnifying fishing effort.

UNOFFICIAL limits 2: the difference between both limits reached up to 4.4 NM in the Blue Hole area. Photo. Click on the image to enlarge it.

Conclusion

The analysis of information provided by fishing activity monitoring platforms must not be carried out while ignoring the context. Some trends shown by analysts—such as, for example, changes in service ports of distant-water fishing fleets due to more rigorous or, conversely, more lax controls—when analyzed in context may simply be due to annual variations in fishery resources or commercial or logistical reasons. This happened with the collapse of the Tsakos Dock in the port of Montevideo in December 2022, which forced fishing vessels operating at Mile 201 to initially seek an alternative for repairs in ports in southern Brazil.

Foto: PNA

With good use, interpretation, and contextualization of information, the combination of data from these platforms and the work of experts and researchers for the monitoring and control of fishing activities is enormously useful. But if this combination does not work, the distortion relative to what occurs in reality can be such that it hinders the search for solutions, especially when such information is taken up by decision-making levels. This article is not intended to criticize monitoring platforms or the work of experts or researchers; on the contrary, the contribution of information and work is greatly valued. It simply attempts to account for the risks represented by working with this technology out of context and without the necessary expertise.

Synthetic aperture radar satellite images provided by CONAE for the detection of vessels that turn off their AIS.

Sergio Almada: Bachelor in Maritime Security – Bachelor and Professor in Geography – Diploma in Policies for the Sustainable Future of the Sea – Coordinator of the Interdisciplinary Team for the Control of Maritime Spaces and their Resources (EICEMAR) – Consultant for the Global Program against Maritime Crime of the United Nations Office on Drugs and Crime (UNODC)

This article represents the criteria of the person signing it. Published opinion pieces do not necessarily reflect the editorial position of this media outlet. Fish Info & Services is an independent medium, open to the opinion of its readers. If you wish to publish, contact us: [email protected]

[email protected]
www.seafood.media

Print