Facing a 1.3-million-km² ocean territory patrolled by only a handful of boats, Seychelles piloted AI-equipped long-range drones in 2018 that autonomously classify vessels for illegal fishing, then formally procured two operational units for its Air Force and Coast Guard in 2021.
>60%
Decline in main fish stocks over three decades (pre-programme)
1.3-1.4 million km²
Seychelles Exclusive Economic Zone area
$150,000 (National Geographic Society Marine Protection Prize)
Initial pilot funding (2018) (2018)
under SCR100,000 (~$7,000)
Drone procurement cost (2021, two units) (2021)
~10,000 km²
Surveillance coverage per drone sortie
12 nautical miles
Some drone models' operational range limit
Details
Maturity
Established
Promoter
Seychelles Fisheries Authority (with Seychelles Coast Guard and Air Force; original pilot with ATLAN Space, Trygg Mat Tracking and GRID-Arendal)
Seychelles' Exclusive Economic Zone spans roughly 1.3–1.4 million km² — one of the largest in the world relative to its landmass — yet is patrolled by only a small number of coast guard vessels. Illegal, unreported and unregulated (IUU) fishing has been linked to a more than 60% decline in the country's main fish stocks over the preceding three decades, threatening both food security and tourism revenue.
Activities
In 2018, the Seychelles Fisheries Authority partnered with Moroccan drone company ATLAN Space, Trygg Mat Tracking and GRID-Arendal on 'Project FishGuard', funded by a $150,000 National Geographic Society Marine Protection Prize. Long-range autonomous drones used onboard AI to classify vessels and detect suspicious activity in near-real time, relaying vessel location and identification data via satellite link without requiring a human pilot or analyst to monitor every flight. By 2021, the Seychelles Fisheries Authority had moved from pilot to institutional operation, procuring two AI-enabled DJI Mavic Enterprise drones for under SCR100,000 (about $7,000), flown by Seychelles Air Force personnel trained alongside the Coast Guard.
Results
Each drone unit can cover roughly 10,000 km² per sortie, extending surveillance reach well beyond what patrol boats alone can achieve. Independent reporting notes real constraints: Africa Defense Forum found some drone models limited to a 12-nautical-mile range and vulnerable to tropical humidity, and flagged that not all jurisdictions accept drone-collected footage as court evidence for prosecutions.
Conclusions
No public source reports arrest or conviction figures directly attributable to the drones, and the fleet remains small relative to the size of the EEZ it covers.
Implementation
Indicative cost
Low (< €50k) — Initial 2018 pilot funded by a $150,000 National Geographic Society Marine Protection Prize; 2021 procurement of two drones cost under SCR100,000 (~$7,000).
Time to results
Long (> 3 years) — Pilot launched 2018; moved to institutional operation by 2021; ongoing as of 2026.
Staffing & skills
Seychelles Air Force personnel (drone pilots), trained alongside the Coast Guard, Seychelles Fisheries Authority operational staff, Original 2018 pilot delivery by ATLAN Space, Trygg Mat Tracking and GRID-Arendal
Conditions for success
Autonomous, AI-based vessel classification reduces the need for a human analyst to monitor every flight
Satellite relay of vessel location/ID data in near-real time
Transition from a grant-funded pilot (2018) to institutional procurement and operation (2021) by the Fisheries Authority itself
Common failure modes
Some drone models limited to ~12-nautical-mile range and vulnerable to tropical humidity (Africa Defense Forum)
Not all jurisdictions accept drone-collected footage as court evidence for prosecutions
No public source reports arrest or conviction figures directly attributable to the drones
Small fleet (2 drones) relative to the 1.3-1.4 million km² EEZ it covers
Where it fits
Governance type
national fisheries authority with military/coast guard support
Scale
national EEZ, ~1.3-1.4 million km², small island state
Income level
upper-middle/high-income (Seychelles)
Commonly funded by
Philanthropic / foundation fundingNational / regional programmes
Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.
Do you run this practice?
Claim it —
verified implementers get a public contact pathway and can propose corrections.
Data sources
Where this practice's information was retrieved from, and when.
MLIT runs a national catalog certifying AI and sensor-based bridge/tunnel inspection technologies against published performance criteria. Listings grew from 12 …