A patrol boat off the coast of Gabon intercepted three trawlers in July 2026, inside the Grand Sud Reserve. This is one of west and central Africa’s largest marine reserves, where industrial trawling is banned. None of the trawlers held a valid licence. The patrol found them because analysts at the non-profit Global Fishing Watch had studied satellite data and worked out where to look.
Most illegal fishing is not caught at all. In 2024, researchers at Global Fishing Watch found that between 72% and 76% of the world’s industrial fishing vessels never appear in public tracking. Most of that activity is concentrated around south and south-east Asia and Africa.
Africa has about 30,500km of coastline shared among 38 coastal states, and illegal fishing costs African economies an estimated US$11 billion to US$13 billion a year.
The reason illegal fishing continues is not complicated. Public tracking depends on vessels reporting their own positions through the automatic identification system, a transponder designed to stop ships colliding. A crew that does not want to be seen switches it off.
But African governments are still leaning on the system of self-reporting. In June 2026, at the Our Ocean Conference in Mombasa, 15 governments signed a declaration promising to open up information about who owns fishing vessels and where they operate. Seven were African, Somalia among them.
There is another way to track illegal fishing, however: from satellites that look for vessels directly.
I am a PhD candidate in aerospace engineering at the University of Florida, where I design spacecraft systems and the software that coordinates them. My supervisor, Norman Fitz-Coy, and I designed a tracking system for the Somali Basin, the stretch of the western Indian Ocean off the Horn of Africa between Somalia and Kenya. It is one of the richest tuna grounds in the world and among the least patrolled.
We concluded that the watching is the easy part. A system that could keep these waters under near-constant observation is affordable and can be built with technology that exists today. But it has to be designed for the Equator instead of borrowed from designs built for other latitudes. Also, even a perfect version runs into a limit that has nothing to do with satellites.
That limit is enforcement. A satellite can show you a vessel. It cannot board one, and it cannot bring a case to court. Overcoming that takes the things no spacecraft supplies: neighbouring states willing to share the cost and the patrols, and the political will to act on what they are shown.
Key design elements
Orbit
Almost every Earth observation satellite flies an orbit that passes close to the poles and crosses the Equator at the same local time each day. It spends much of each circuit above the Arctic, the Antarctic, and the near-empty latitudes between. For a country sitting on the Equator, such as Kenya, Somalia or Gabon, most of those laps are wasted.
We modelled a shell of 24 satellites tilted at just 20 degrees to the Equator. Over the Somali Basin it returns roughly every 1.1 hours, against 3.4 hours for the same 24 satellites flown over the poles. Put plainly, satellites hugging the Equator see the basin about three times more often. Commercial radar operators have begun flying tilted orbits at 40 and 45 degrees for this reason, but none fly low enough to concentrate on these waters.
Sensors
Finding a boat and identifying it are two different jobs. Radar picks out a vessel as a bright dot across a wide strip of ocean. Telling whether that dot is a licensed trawler or a 7 metre skiff needs an image sharp enough to resolve detail well under a metre, and no single sensor does both. A wide sensor finds candidates and a sharp one is pointed at them afterwards, so the system must decide in orbit, within seconds, which dots deserve a second look.
Power and data
Radar is power hungry. A 100kg radar satellite carries enough stored energy to image for about 4.5 minutes per orbit and spends the remaining 90 or so minutes charging. Those few minutes still generate roughly 690GB a day, while the ground stations serving the region can pull down between 120GB and 360GB. There are not many of them. The main ones are the Italian-run Broglio Space Centre at Malindi on the Kenyan coast, and the South African National Space Agency’s antennas at Hartebeesthoek, some 4,000km to the south.
So, the satellite must sort its own pictures before sending them, keeping what might be a vessel and discarding the rest. That changes what it has to be. Not a camera shipping pictures home for people to examine, but something that recognises a boat by itself, in orbit, with nobody watching.
Cost
The system we costed has all of those constraints designed into it, the tilted orbit, the two kinds of sensor, the onboard sorting and the thin downlink. It comes to about US$347 million over five years, or roughly US$25 for every confirmed detection. Against US$11 billion in annual losses from illegal fishing, cost is not the obstacle.
The harder part is getting it used
A constellation is only worth building if governments act on what it shows, and some countries have already shown what that commitment looks like using the self-reporting data they have now. Indonesia, Peru, Chile, and Costa Rica publish their national vessel tracking data, turning an enforcement feed that was private into something anyone can check. East Africa has its own version.
FISH-i Africa, a task force of eight western Indian Ocean states including Somalia and Kenya, has since 2012 pooled intelligence on suspect vessels across borders. It has worked over 35 investigations, caught vessels sailing under forged licences and false identities, and had illegal operators fined and brought to court, largely by sharing information its members already held.
So, what would it take for African governments to buy into a constellation? Mostly, not paying for it alone. The US$347 million works out at about US$69 million a year. Split eight ways across the FISH-i Africa states, that is roughly US$9 million a country per year for continuous coverage of waters they already share. The expensive way to buy this capability is one nation at a time.
What needs to happen next
Our study set out six strategies for using space and airborne systems to support African maritime security. Five are ordinary engineering, covering the constellation itself, fusing radar with other sensors, the ground computing to handle what comes down, smaller and better sensors, and aircraft or drones to fill the gaps between satellite passes.
We rejected the sixth ourselves. It was a concept for microsatellites carrying high-powered lasers to disable a pirate vessel’s engine or communications from orbit, and it failed on the physics before anyone reached the legal objections.
The finding that surprised us most was where the system finally breaks. It is not the cameras. It is the courts. A vessel that has been seen, identified and reported still has to be intercepted, boarded, charged and tried. Where that does not happen, the detection produces a cost and no security.
In Gabon, the patrol was only half the story. The trawlers’ owner later settled with the authorities, and getting there took inspectors, prosecutors, courts and the legal authority to board a foreign vessel. Those are people, institutions, and laws.
Africa can build the eyes. The orbit is understood, the sensors exist and the price is modest next to the losses. The Mombasa declaration is a reasonable place to begin, but it rests on a system three quarters of the fleet has already stepped outside. Seeing the outsiders is an engineering problem with an answer. What happens next is the harder one.
- is a Ph.D. Candidate in Aerospace Engineering, University of Florida
- This article first appeared on The Conversation
