Turning subsea fiber into a tripwire for narco-submersibles.
Hydroacoustic classification of semi-submersible and fully submersible vessels from distributed acoustic sensing on existing telecommunication cable, for transnational interdiction and maritime domain awareness.
Concept briefing · UnclassifiedIllustrative simulation. Waveforms, signal-to-noise values and confidence levels are notional: no absolute strain levels or detection performance for these target classes have been published. Real seafloor DAS data are dominated by ocean-wave noise at roughly 0.01–0.26 Hz, shown here as the background swell.
Established sensing, new target set
Distributed Acoustic Sensing (DAS) over subsea telecommunication fiber is established science. Seismologists have used submarine telecommunication cables to record earthquakes, microseisms, ocean waves and ship traffic, turning tens of kilometres of fiber into thousands of strain sensors.
The operational pivot. Existing processing chains are tuned for large, loud or slow phenomena. Low-profile smuggling hulls, built from fiberglass, wood, Kevlar or carbon fibre with signature-reduction measures, sit beneath those thresholds. ExNovo SSD moves the problem from broadband seismic monitoring to micro-signature isolation: separating the acoustic and hydrodynamic fingerprints of these craft from natural ocean noise.
Two target classes, two physical channels
Self-propelled semi-submersible (SPSS)
Rides low at the surface boundary (typically about 18 in, or 0.46 m, of freeboard according to a 2011 Naval Postgraduate School thesis) on commercial diesel engines. Uncrewed variants, first captured at sea by Colombia in July 2025, keep this low profile and add satellite links and cameras.
- Acoustic channelExpected harmonic lines between roughly 10 and 200 Hz from engine firing rate, shaft rate and propeller blade rate. Uncrewed variants still carry an engine, so the same channel applies; their source levels are not publicly documented.
- Hydrodynamic channelSurface wake only; negligible pressure coupling into the seabed.
Fully submersible vessel (FSV)
Rare. A sealed hull that dives fully and runs on batteries while submerged. Diesel-electric types (Ecuador, 2010) must surface or snorkel to recharge; a battery-electric craft was seized in Colombia in 2020.
- Acoustic channelExpected to be the quietest class while on batteries: no exhaust and little cavitation at low speed. Diesel-electric types are loud again when recharging at the surface.
- Hydrodynamic channelPublished work shows seabed DAS detecting the low-frequency pressure field of moving ships and diving whales, with signals that decay quickly with distance. Applying this to a small submerged hull is a hypothesis: the expected range is short and is to be characterised in Phase I.
| Metric | Ambient background | SPSS · diesel | FSV · electric |
|---|---|---|---|
| Primary channel | Stochastic wave and microseism noise | Acoustic spectral lines | Quasi-static seabed pressure strain |
| Frequency window | Broadband | ≈10–200 Hz harmonics | <1 Hz |
| Published analogue | Ocean-wave noise dominated roughly 0.01–0.26 Hz on a Belgian cable | Vessel detection F1 above 90% within a 1 km threshold on a buried cable (Belgium) | Moving ships and diving whales detected through low-frequency pressure fields; signals decay quickly with distance (Svalbard) |
| Program objective | False-alarm baseline | Detection and false-alarm rates set from Phase I measurements | Detection range for small hulls measured in Phase I; no published figure exists |
| Deep-water expectation | Stable background | Degraded: low SNR after spreading and absorption | Not expected to be observable: pressure footprint decays before reaching the seabed |
No detection probability for these target classes has been published. Phase I measures noise floor and signal levels; Phase III validates detection and false-alarm rates with proxy vessels.
Open-source field record
| Date | Where | Type | Reported details |
|---|---|---|---|
| 2 Jul 2010 | Near the Colombian border, Ecuador | Fully submersible, diesel-electric, twin screw | DEA: about 30 m long, with a conning tower, periscope and air conditioning. NPR adds 249 lead-acid batteries, up to 18 h submerged, maximum dive about 19 m (62 ft), 6,800 nmi range and about 9 t capacity, and gives the length as 74 ft (22.6 m). DEA · NPR |
| Feb 2011 | Timbiquí, Cauca, Colombia | Fully submersible, twin diesel | AP: 99 ft (30 m) fiberglass hull; dives to about 9 ft (3 m); about 8 t capacity; crew capacity six. Fox News (AP) |
| Jul 2017 | San Juan and Baudó river area, Colombia | Battery-powered | Infobae: more than 100 batteries, no exhaust, being outfitted for more than 4 t. RUSI also records a 2017 battery-propelled, fully submersible seizure. Breitbart (citing Infobae) · RUSI |
| 5 Nov 2020 | Chocó, Colombia | Electric | Colombian Navy estimate: up to 6 t. Hydroplanes suggest a fully submersible hull, an analyst's inference from images. Reports differ: about 20 m long and 3,000+ km range (MercoPress); about 12 h submerged, or 32 nmi at 3 kn, on 10 t of batteries (USNI). H. I. Sutton · USNI News · MercoPress |
| Jul 2022 | Spain | Uncrewed, remotely operated | Three seized, two under construction and one nearly finished; up to 200 kg each. CBS News |
| Announced 2 Jul 2025 | Caribbean coast of Colombia | Uncrewed semi-submersible | Empty, on a reported test run; up to 1.5 t capacity; satellite antenna (Starlink inferred from images), cameras and GPS; an analyst quoted by Naval News says it is not truly submersible. Naval News · CBS News |
Figures are as reported by the cited outlets and are not official specifications. Where sources disagree, both figures are shown.
Mathematics & wave mechanics
Three relationships govern what the interrogator can see.
Axial strain over a gauge length L shifts the Rayleigh-backscatter phase:
Received level at the cable from an SPSS source, with spherical spreading to the transition range rt and cylindrical spreading beyond it in a shallow waveguide:
For a hull moving at speed U with velocity potential Φ (seabed-fixed frame, so the hull moves through the field), the dynamic pressure at the seabed follows from unsteady Bernoulli:
Telemetry frame standard
Each interrogator emits time-stamped feature frames for the inference layer. Schema (abridged) and an illustrative sample frame; the values are not measurements:
{
"$schema": "https://json-schema.org/draft/2020-12/schema",
"title": "ExNovoSsdTelemetryFrame",
"type": "object",
"required": ["timestamp", "sensor",
"acoustic", "hydrodynamic", "inference"],
"properties": {
"timestamp": { "type": "string", "format": "date-time" },
"sensor": { "type": "object", "properties": {
"interrogator_id": { "type": "string" },
"channel": { "type": "integer" },
"sampling_rate_hz": { "type": "number" } } },
"acoustic": { "type": "object", "properties": {
"rms_phase_rad": { "type": "number" },
"psd_peaks_hz": { "type": "array",
"items": { "type": "number" } } } },
"hydrodynamic": { "type": "object", "properties": {
"quasi_static_snr_db": { "type": "number" },
"event_duration_s": { "type": "number" } } },
"inference": { "type": "object", "properties": {
"predicted_class": { "enum": ["BACKGROUND",
"SPSS_DIESEL", "FSV_ELECTRIC", "UNKNOWN"] },
"confidence": { "type": "number",
"minimum": 0, "maximum": 1 } } }
}
}
{
"timestamp": "2026-11-09T21:45:00Z",
"sensor": {
"interrogator_id": "EXN-DAS-04B",
"channel": 1184,
"sampling_rate_hz": 2000
},
"acoustic": {
"rms_phase_rad": 0.0842,
"psd_peaks_hz": [12.4, 24.8, 49.6]
},
"hydrodynamic": {
"quasi_static_snr_db": 9.5,
"event_duration_s": 14.5
},
"inference": {
"predicted_class": "FSV_ELECTRIC",
"confidence": 0.915
}
}
Milestones & phase budget
A 24-month program in three gated phases.
Signal cataloging and hardware deployment: narrow-linewidth phase-OTDR interrogators installed on shallow coastal cable landings.
Filtering and classifier development: separating vessel signatures from biological, weather and shipping noise.
Field validation with proxy vessels to measure detection and false-alarm rates, and delivery of API alerting.
References
- Hartog, A. H. (2017). An Introduction to Distributed Optical Fibre Sensors. CRC Press.
- Lindsey, N. J., Dawe, T. C., & Ajo-Franklin, J. B. (2019). Illuminating seafloor faults and ocean dynamics with dark fiber distributed acoustic sensing. Science, 366(6469), 1103–1107.
- Williams, E. F., et al. (2019). Distributed sensing of microseisms and teleseisms with submarine dark fibers. Nature Communications, 10, 5778.
- Rørstadbotnen, R. A., & Landrø, M. (2026). Detecting silent whales using seabed fiber-optic cables. PNAS, 123(26), e2603077123. A correction notice followed (PNAS 123(41), doi:10.1073/pnas.2632392123).
- Ramirez-Torres, E. E., et al. (2025). Vessel detection and localization using distributed acoustic sensing in submarine optical fiber cables. arXiv:2509.11614.
- Watkins, L. J. (2011). Self-Propelled Semi-Submersibles: The Next Great Threat to Regional Security and Stability. Master's thesis, Naval Postgraduate School.
- Urick, R. J. (1983). Principles of Underwater Sound (3rd ed.). McGraw-Hill.