Sensors
- HawkEye 360 RF constellation: Clusters of three small satellites flying in formation detect and geolocate emitters across 144 MHz to 15 GHz using time-difference-of-arrival and frequency-difference-of-arrival. Geolocation accuracy is typically 500 m to 2 km depending on geometry and signal duration. Revisit over a fixed offshore location is several times per day at mid-latitudes, improving toward the poles.
- Spire SDR (software-defined radio) constellation: Spire operates more than 100 LEO satellites carrying software-defined radio payloads that can be retasked to specific frequency bands. Useful for monitoring VHF marine band (156–174 MHz), AIS (161.975 and 162.025 MHz), and platform-specific licensed frequencies. Revisit is frequent but dwell time per pass is short, typically under 10 minutes per satellite.
- Spire AIS (space-based AIS): Decodes Class A and Class B AIS transponder messages from vessels attending or near fixed infrastructure. Absence of expected supply-vessel traffic around a normally active platform is a secondary indicator of operational disruption. Global coverage with sub-hourly revisit at most latitudes.
- Sentinel-1 SAR (C-band, 5.405 GHz): Provides structural and surface confirmation of platform presence and condition. Interferometric coherence over a fixed steel structure is very high; a sudden coherence drop can indicate structural change or flooding. Ground range detected mode achieves 20 m resolution with 6-day repeat at mid-latitudes (12-day for a single satellite). Free and open data from ESA Copernicus.
What a working platform sounds like from orbit
A staffed offshore oil platform is one of the noisier fixed points on the radio spectrum. It runs licensed VHF voice channels for deck operations, a dedicated VSAT terminal for crew communications and SCADA telemetry, X-band and S-band navigation radars rotating continuously, and an AIS transponder reporting its position as a fixed aid to navigation. Some platforms also carry licensed microwave links to shore or to neighbouring structures. That aggregate emission pattern is stable and repeatable across years.
Spaceborne RF sensors, particularly HawkEye 360's formation-flying clusters, detect and geolocate these emitters without any cooperation from the platform. The physics is straightforward: each satellite in a cluster measures the time and frequency at which the same signal arrives, and the differences in those measurements constrain the emitter's location on the ground. For a fixed structure, the geolocation simply needs to be consistent with the known coordinates across successive passes. Deviation is the finding.
Silence, substitution and the anomalies worth investigating
Three RF anomaly classes matter for fixed offshore infrastructure. The first is complete silence: a platform that has stopped transmitting on all expected channels simultaneously. This can indicate an emergency evacuation, a catastrophic power failure, or deliberate RF suppression ahead of an illicit operation. It is distinct from a single failed transmitter, which leaves the remaining emissions intact.
The second class is frequency substitution: the expected licensed emissions are absent but new, unlicensed transmissions appear at or near the same coordinates. This pattern is associated with vessels using a fixed structure as a relay point for communications they do not want attributed to a moving ship, a tactic documented in sanctions-evasion contexts.
The third is timing anomaly: emissions are present but their duty cycle or modulation changes. A VSAT terminal that normally transmits continuously but switches to burst-mode at irregular intervals may indicate a change in operational tempo, a crew reduction, or equipment fault. None of these inferences is conclusive from RF alone, which is precisely why fusing RF observation with SAR and optical imagery matters.
What a floating roof gives away, and what SAR adds
Sentinel-1 SAR does not detect RF emissions. What it does is confirm that the physical structure is present, intact, and producing a coherent radar return consistent with its known geometry. A steel platform produces a bright, stable point target in SAR imagery. Persistent scatterer interferometry over a fixed structure can detect millimetre-scale deformation over months, which is relevant for ageing infrastructure and for subsidence above subsea completions.
The combination is more informative than either sensor alone. If SAR shows the structure is physically present and undamaged, but RF monitoring shows all emissions have ceased, the silence is more alarming than it would be if the structure had also disappeared from radar. Conversely, if RF shows normal emissions but SAR shows a changed backscatter signature around the base of a jacket, the structural concern overrides the reassurance from the radio.
For subsea cable landing stations, which are onshore or nearshore structures rather than open-ocean platforms, SAR is less useful but RF monitoring of the cable's associated communications equipment remains applicable. Cable faults are sometimes preceded by changes in the modulation or power level of the shore-end terminal equipment, though detecting these from orbit requires tasking Spire SDR payloads to the relevant licensed microwave or satellite backhaul frequencies.
Honest limits of the method
Geolocation accuracy from HawkEye 360 is not fine enough to distinguish between two transmitters on the same large platform. For a structure that spans 100 m or more, the RF observation tells you the platform is active; it does not tell you which specific antenna or deck is the source. That matters when trying to distinguish a legitimate crew communication from an unauthorised transmitter co-located on the same structure.
Revisit is the other constraint. A HawkEye 360 pass over a fixed North Sea location might occur four to eight times per day, depending on orbital geometry at that latitude. A brief, opportunistic transmission lasting only a few minutes may fall entirely between passes. Spire's larger constellation helps fill gaps, but neither network offers continuous coverage of a single point. Events shorter than roughly 10 to 15 minutes carry a meaningful probability of being missed in any given 24-hour window.
Cloud cover does not affect RF or SAR collection, which is one reason this sensor combination is preferred over optical-only workflows for offshore monitoring. Optical imagery from Sentinel-2 or commercial providers is still useful for confirming vessel presence and flare status, but it is not load-bearing for the RF anomaly use case.
Building a baseline and setting thresholds
The analytic value of RF monitoring over fixed infrastructure is almost entirely relative. A single observation that a platform is transmitting on a given frequency tells you little. A 90-day archive of pass-by-pass observations that establishes the normal emission profile, and then flags deviations from it, tells you a great deal.
Baseline construction requires collecting RF detections across enough passes to characterise the platform's typical active frequencies, transmission schedules, and signal strengths. For a well-staffed production platform, this baseline stabilises within two to four weeks of consistent collection. For a normally unattended wellhead or cable repeater station, the baseline is simpler: expected silence punctuated by known maintenance windows.
Satellize structures this kind of persistent monitoring as a scheduled analytics layer, drawing on licensed commercial RF data and open SAR archives. The approach is similar in principle to the change-detection logic used in the Tonga crop-estimation programme, where the analytic value comes from comparing current observations against a stable historical reference. Thresholds for alerting are set in consultation with the client, because a national oil company and a maritime insurer have very different tolerances for false positives.
What the output looks like in practice
The deliverable for an offshore RF monitoring programme is not a raw spectrum recording. It is a structured alert feed, keyed to specific infrastructure assets, that flags pass-by-pass deviations from the established baseline. Each alert carries the detection timestamp, the satellite pass geometry, the observed frequency or silence condition, a confidence score based on how many satellites in the formation contributed to the geolocation, and a recommended cross-check action.
For clients managing a portfolio of platforms across a basin, the feed integrates with existing maritime domain awareness tools as a GIS layer or API endpoint. For clients focused on a single critical asset, a daily digest report is often more useful than a real-time feed. The choice depends on operational tempo, not on any inherent limitation of the data.
Typical figures
| RF geolocation accuracy (HawkEye 360) | Typically 500 m to 2 km, dependent on signal duration, geometry and frequency |
| Frequency coverage (HawkEye 360) | 144 MHz to 15 GHz (covers VHF marine, AIS, X/S-band radar, VSAT uplinks) |
| RF revisit over a fixed offshore point | 4 to 8 passes per day at mid-latitudes (HawkEye 360); higher with Spire SDR augmentation |
| Minimum detectable signal | Dependent on emitter EIRP and frequency; continuous VSAT and radar emitters reliably detected; brief low-power bursts may be missed |
| SAR spatial resolution (Sentinel-1 IW mode) | 20 m ground range detected; 5 × 20 m in single-look complex |
| SAR revisit (Sentinel-1) | 6 days at mid-latitudes with two-satellite constellation; 12 days with one |
| SAR archive depth (Sentinel-1) | From April 2014 (Sentinel-1A launch); free and open via Copernicus Data Space |
| AIS message latency (Spire) | Typically under 60 minutes from transmission to delivery via API |
| RF archive depth (HawkEye 360) | Commercial archive from 2019 onwards; coverage density increases with constellation growth |
| Delivery formats | GeoJSON alert feed, GIS layer (GeoPackage or Shapefile), API endpoint, PDF digest report |
Analytics Satellize can run
| Platform RF baseline profile | Statistical characterisation of observed emission frequencies, duty cycles and signal-present/absent patterns across 30 to 90 days of RF passes | Asset-specific baseline report with frequency inventory and normal operating envelope, delivered as PDF and structured JSON |
| Emission silence alert | Pass-by-pass comparison of observed RF detections against baseline; alert triggered when all expected emission categories are absent across two or more consecutive passes | Real-time alert via API or email, with pass geometry, timestamp and confidence score |
| Unlicensed or anomalous frequency detection | Frequency-domain comparison of detected emissions against the platform's licensed spectrum assignments; flagging of signals outside the expected set | Incident report with detected frequency, geolocation, signal characteristics and recommended follow-up action |
| SAR structural change detection | Sentinel-1 coherence change analysis and backscatter intensity time series over fixed platform coordinates; persistent scatterer methods for deformation where archive depth allows | Monthly SAR change report as GeoTIFF and PDF, with flagged epochs and annotated imagery |
| Vessel attendance cross-check | Space-based AIS track aggregation around platform coordinates; comparison of supply-vessel visit frequency against client-supplied operational schedule | Weekly vessel attendance summary as CSV and GIS layer, with anomaly flags for unexpected absences or unscheduled visits |
| Multi-sensor anomaly fusion report | Temporal alignment of RF alert flags, SAR change detections and AIS attendance gaps to identify correlated anomalies across sensor types; analyst-reviewed synthesis | Quarterly or on-demand fusion report as PDF, with timeline of correlated events and recommended investigative actions |
Who does the work
We can get this done for you. Satellize runs its own analyst desk and a strong science team. You do not buy a data feed and work out what it means; our people source the imagery, run the analysis described on this page, and hand you the answer with its confidence limits stated. Discuss this requirement.