Ports combine long dark waterlines, high-value cargo and constant vessel movement. EO/IR drone patrol gives harbour security a moving vantage point that fixed cameras cannot match — with thermal for detection over black water and zoom for identification at anchorage distance. Water is the best thermal background there is, right up until it is warm.
- 30xLX-9B optical zoom
- 0.8 W x2LX-9B NIR illuminators
- 2,000 mLaser ranging class
- Salt airFirst-order design constraint
Key takeaways
- Water renders cold and uniform in thermal, so vessels and small craft stand out sharply. That advantage collapses for a person in warm water.
- Hull names and container codes are zoom-channel work. Thermal will never read markings.
- Salt atmosphere makes ingress protection and corrosion care first-order concerns, ahead of optical specification.
- Coordinates matter more here than almost anywhere: a patrol boat needs a position, not a description of a dark waterline.
On this page
Waterfront patrol at night
Water renders cold and uniform in thermal, so vessels, swimmers and small craft stand out sharply. A harbour basin at night is close to an ideal thermal scene: a large, flat, thermally uniform background against which anything warm is unambiguous.
Vessels are the easiest targets on this mission. An engine-warm hull presents a large, strongly contrasting signature detectable at kilometres, and even a cold hull sits distinctly against water. Small craft running dark are detected on thermal long before they are visible.
The exception is people in the water, and it is a large exception. Detection ranges for person-in-water scenarios track the published person DRI figures, but they degrade sharply with water temperature because a wet head approaches water temperature within minutes. In warm water the contrast can fall below one kelvin. The rescue side is covered on the SAR page, and open-water specifics on maritime SAR.
| Target | Thermal detection | Zoom identification | Notes |
|---|---|---|---|
| Vessel, engine running | 2,000–4,000 m | 400–900 m | Engine and exhaust dominate |
| Vessel, cold hull | 1,000–2,000 m | 400–800 m | Hull-vs-water contrast |
| Small craft / RIB, running | 1,500–3,000 m | 300–600 m | Wake often visible first |
| Small craft, drifting dark | 600–1,200 m | 250–500 m | The security-relevant case |
| Person on a jetty or deck | 800–1,400 m | 200–400 m | Standard land-type contrast |
| Person in water, cold (5–12 C) | 300–600 m | 100–200 m | Wet head only |
| Person in water, warm (over 26 C) | 100–250 m | under 100 m | Contrast near zero |
| Container markings | — | 150–400 m | Zoom only; thermal cannot read text |
| Hull name / IMO number | — | 200–600 m | Zoom plus NIR at night |
Wave clutter defeats automatic detection over water. Reflected sky on moving wave facets produces a continuous stream of false positives that most onboard trackers cannot filter. Automation works well for vessels and poorly for swimmers.
Vessel and yard identification
Hull names and container codes are zoom-channel work: 30x optical class reads markings from standoff, and NIR laser illumination extends identification into unlit berths. AI tracking holds a moving small craft while the operator coordinates response.
This is the operational core of port security. Detection is rarely the hard part in a harbour — the water makes that easy. The hard part is answering which vessel, which container, which berth, and doing so without closing the distance or lighting the subject.
Container yard work adds a geometric problem the waterline does not have. Stacked containers create deep shadow and steep viewing angles, and a code readable from directly above may be unreadable from an oblique orbit. Plan orbit geometry around the stack orientation rather than the fence line.
Salt atmosphere: the constraint that outranks optics
Salt atmosphere makes ingress protection and corrosion care first-order concerns. This is not a footnote on a port programme — it is the leading cause of payload attrition.
Salt aerosol penetrates further than dust and attacks differently. It creeps into connector shells, promotes galvanic corrosion at dissimilar-metal interfaces, and leaves a hygroscopic film on optics that attracts moisture long after the flight. A payload rated for rain is not automatically suited to daily marine exposure.
Three practical measures. Specify ingress protection for the full assembly including mated connectors — see what IP ratings really promise. Rinse and dry airframe and payload after marine sorties as routine, not as maintenance. And expect shorter service intervals on gimbal mechanics and connectors than an inland deployment would suggest.
Handing coordinates to a patrol boat
A contact on a dark waterline is meaningless to a patrol boat without a position. Featureless water offers no landmarks to describe against, which makes this mission more coordinate-dependent than most.
Laser ranging computes the target position directly, and the resulting coordinate can be passed straight to the responding vessel. Report an accuracy radius alongside it: over water the observation is usually oblique and at long range, which is where angular error contributes most — the terms are in the geolocation error budget.
AI tracking matters operationally here because the target moves and the aircraft must reposition. Holding a small craft automatically while the operator talks to the patrol boat is the difference between a coordinated interception and a lost contact.
Where port programmes go wrong
- Assuming person-in-water detection matches land figures. It does not, and in warm water it collapses entirely.
- Relying on automatic detection over water. Wave clutter generates continuous false positives for swimmer-sized targets.
- Under-specifying for salt exposure. Connectors and gimbal bearings fail long before optics do.
- Searching into sun or moon glint. Specular reflection off wave faces hides exactly the small warm targets you want.
- Orbit geometry planned around the fence, not the stacks. Container codes need a viewing angle, not just a distance.
- No coordinate hand-off. Descriptions do not work on featureless water.
Recommended UAVThermal payloads
LX-9B — thermal with 30x zoom and dual NIR laser illumination for unlit waterline identification, where berths and basins have no ambient light.
OP-125A — kilometre-class person detection with 2,000 m laser ranging and AI tracking for wide basin coverage and moving contacts.
LX-8RC — long-reach hybrid-zoom night-vision payload for anchorage-distance observation and vessel identification at range.
Related reading
Technology: NIR laser illumination and starlight and low-light imaging.
Field practice: real detection data from six payloads, maritime SAR and overwater operations and IP ratings explained.
Adjacent missions: maritime SAR and coastguard and border and perimeter surveillance.
FAQ
Can thermal find a person in the water at night?
Yes in cold water, poorly in warm water. A person in the water exposes a wet head that approaches water temperature within minutes, so in 5 to 12 C water detection at 300 to 600 m is realistic while above 26 C contrast can fall below one kelvin and range collapses to a few hundred metres at best.
Can thermal read a hull name or container code?
No. Thermal renders heat, not markings. Text is zoom-channel work — 30x optical class reads hull names and container codes from standoff, with NIR laser illumination extending that into unlit berths at night.
What matters most for a port deployment?
Salt-atmosphere durability, ahead of optical specification. Salt aerosol penetrates connectors, promotes galvanic corrosion and leaves a hygroscopic film on optics. Specify ingress protection for the full assembly including mated connectors, and rinse after marine sorties as routine.
Does automatic detection work over water?
For vessels, yes. For swimmer-sized targets, not reliably — reflected sky on moving wave facets produces continuous false positives that most onboard trackers cannot filter. Person-in-water search still depends on a trained operator.