A radio protocol designed for harbours, received from orbit
The Automatic Identification System was designed in the 1990s to stop ships colliding in confined coastal waters. It was not designed for space. Vessels broadcast a VHF self-organising time-division multiple access (SOTDMA) packet on 161.975 MHz and 162.025 MHz, announcing position, heading, speed, MMSI identifier and vessel type. A shore station a few tens of kilometres away hears a manageable number of transmissions. A satellite at 500 to 600 km altitude, with a footprint exceeding 4,000 km in diameter, hears thousands simultaneously.
That geometric mismatch is the central engineering problem of space-based AIS. It is also, paradoxically, what makes the sensor so valuable: one small receiver, one low orbit, global coverage within hours. No other maritime sensor achieves that combination at comparable cost. The ITU mandates AIS carriage on all SOLAS-class vessels above 300 gross tonnes on international voyages, plus all passenger ships regardless of size, which means the transmitting population is large and legally compelled to participate.
Collision physics: why dense waters corrupt the signal
SOTDMA works by letting each transponder reserve a time slot in a two-minute, 2,250-slot frame. In coastal waters, neighbouring vessels negotiate slot assignments and collisions are rare. From orbit the receiver sees every vessel within the footprint trying to use the same 2,250 slots. When two transmissions overlap at the receiver, both are typically lost. This is packet collision, and it scales badly: detection probability drops from near-100 % in sparse ocean to below 50 % in the Strait of Malacca or the English Channel during peak traffic, according to published analyses from ESA's NORAIS experiment and subsequent studies of the EXACTVIEW and ORBCOMM constellations.
Several techniques reduce the damage. Parallel multi-channel receivers can listen to both AIS channels and additional out-of-band re-use frequencies simultaneously. Wide-band software-defined radio (SDR) receivers digitise a broader slice of spectrum and apply post-processing to resolve overlapping packets by exploiting small frequency offsets between transmitters. Directional or phased-array antennas can narrow the footprint, trading global snap-shot coverage for better collision performance over a chosen region. None of these fully solve the problem in the world's busiest straits; they manage it.
What a constellation actually delivers
A single AIS satellite in a 500 km sun-synchronous orbit revisits any given ocean point roughly every 90 to 100 minutes, though the geometry means high-latitude coverage is better than equatorial. Spire Global operates more than 100 LEO satellites with AIS payloads and publishes revisit statistics showing median global vessel detection latency under 30 minutes for open-ocean traffic. exactEarth (now HawkEye 360-adjacent) and ORBCOMM have published detection rates exceeding 90 % for isolated vessels in low-density ocean areas.
For a national programme, the practical question is whether to fly a sovereign receiver or subscribe to a commercial feed. A single hosted AIS payload, typically 0.5 to 2 kg and drawing 2 to 5 W, can be integrated onto a 6U to 16U cubesat or as a secondary payload on a larger bus. It delivers raw message logs to a national ground station, giving the operator unmediated access to data that commercial aggregators may filter, delay or withhold for commercial reasons. Sovereignty over the raw feed, not just the processed output, is the reason several coastal states have chosen to fly their own receivers rather than rely entirely on third-party services.
The dark-ship problem is a feature, not a flaw, of honest analysis
AIS is a cooperative sensor. A vessel that switches off its transponder disappears from the record entirely. Fishing vessels below the SOLAS threshold are not required to carry AIS at all; many do not. Transponder manipulation, broadcasting false MMSI numbers or spoofed positions, is technically straightforward and documented in open-source maritime security literature. In sanctioned-cargo monitoring, illegal fishing detection and naval surveillance, the ships that matter most are frequently the ones that go quiet.
This is not a reason to dismiss AIS; it is a reason to understand what it does and does not prove. A vessel present in the AIS record is almost certainly where it claims to be: the system is reliable for compliant traffic. A vessel absent from the record proves nothing on its own. Absence becomes meaningful only when correlated with other sensors, such as SAR imagery, optical passes or RF emission data from signal-mapping payloads. AIS is best treated as a cueing and context layer, not a complete picture. Any programme architecture that presents AIS coverage alone as maritime domain awareness is overselling the sensor.
Integration and the honest limits of a hosted payload
A hosted AIS receiver is among the simplest payloads to integrate. It requires a VHF antenna with a clear sky view, a low-noise amplifier, an SDR or dedicated ASIC demodulator, and a data interface to the onboard computer. Power draw is low enough that it can ride as a secondary payload on almost any bus without driving the power budget. Lead time from procurement to flight-ready hardware is typically 6 to 18 months depending on the supplier and whether a heritage design is available.
The limits are real. Frequency congestion in busy straits is not fully solvable from orbit. Revisit from a single satellite is adequate for ocean monitoring but too slow for real-time port surveillance. Raw message volume from a wide-beam receiver over a busy ocean region can run to millions of packets per pass, requiring non-trivial ground processing to deduplicate, validate and geolocate messages. And the sensor tells you nothing about vessels that choose silence. A sovereign programme should budget for ground-segment processing capability and plan from the outset for data fusion with at least one complementary sensor type.
Engineering parameters
| Receive frequencies | AIS CH1: 161.975 MHz, CH2: 162.025 MHz; SDR variants cover broader VHF band |
| Typical payload mass | 0.5 to 2 kg for cubesat-class receivers; up to ~5 kg for phased-array or multi-channel units |
| Power consumption | 2 to 5 W (single-channel); 8 to 20 W for wideband SDR or array variants |
| Footprint diameter at 500 km altitude | Approx. 4,000 to 5,000 km; narrowed to 500 to 1,000 km with directional antenna |
| Packet detection rate (open ocean, low density) | Greater than 90 % for isolated vessels; drops below 50 % in high-density straits |
| Position accuracy of received messages | Vessel-reported GNSS position; typical accuracy 10 m CEP (dependent on vessel equipment) |
| Revisit interval (single satellite, 500 km SSO) | 90 to 100 min per point; high latitudes revisited more frequently |
| Data volume per pass | Tens of thousands to millions of raw packets over busy ocean regions; varies with footprint and traffic density |
| Integration form factor | 6U cubesat secondary payload upward; compatible with most standard bus interfaces (PC/104, SpaceWire, UART) |
| Payload procurement lead time | 6 to 18 months typical for heritage designs; longer for custom phased-array or wideband SDR variants |
One contract, one accountable engineer
Commissioned as one programme, not a stack of contracts: spacecraft, launch, ground segment, mission control, training and handover are priced together. Source-access terms and audit rights are agreed in writing before signature. Request a sovereign AIS architecture review.