Cruise Ships Can Detect Man Overboard in Real Time. Why Isn’t Every Ship Doing It?
🔔 Subscribe to ShipUniverse Weekly →
The Cruise Safety Technology That Took 16 Years to Arrive
The technical problem that frustrated the cruise industry for more than a decade is becoming much harder to call unsolved.
Cameras can watch the ship's perimeter continuously. Thermal imaging can see at night. Radar and computer vision can distinguish trajectories. Software can recognize a fall, generate a position, alert the crew and, in some systems, continue tracking the person after they enter the water.
In 2026, a system completed independent ISO testing aboard an operating cruise ship and subsequently detected a real overboard incident in real time. Yet automatic detection remains far from a universal cruise-ship feature.
The question facing cruise operators in late 2026 is changing. It is no longer simply whether automatic MOB detection works. It is whether the remaining barriers justify keeping it off ships that could technically carry it.
Congress acts
The Cruise Vessel Security and Safety Act creates the U.S. requirement concerning technology capable of capturing images or detecting passengers falling overboard.
MSC demonstrates detection
MSC Meraviglia enters service with intelligent optical and thermal surveillance. MSC reports 97% accuracy after more than 25,000 hours of analysis and testing.
A standard arrives
ISO 21195 establishes an international performance framework for automatic person-overboard detection systems.
Certification crosses the line
Zelim's ZOE completes all three ISO test phases aboard Ambition and receives ISO 21195 certification through Lloyd's Register.
The law contains one word that changed the technology race
U.S. law does not simply say that every covered cruise ship must have an automatic man-overboard alarm. The statutory wording allows technology capable of capturing images of passengers or detecting passengers who fall overboard, to the extent such technology is available.
Conventional CCTV can record the exterior of the vessel without necessarily recognizing an overboard event as it happens.
Automated MOB technology analyzes activity around the vessel and can generate an immediate alarm when its detection criteria indicate a human overboard event.
The regulatory gap is still open
The U.S. Coast Guard's current public FAQ says it is not presently enforcing the CVSSA provision covering image capture and detection of persons falling overboard. The Coast Guard says implementing regulation and policy direction are still required.
That means technological capability and mandatory fleet deployment are not presently moving on the same timetable.
Why false alarms held the industry back
One false alarm is not one harmless notification
An automatic MOB warning reaches a safety-critical operating environment. If crews repeatedly encounter alarms caused by birds, spray, reflections or objects falling from decks, confidence in the system can deteriorate.
Scale that across a 25-ship fleet
Even a seemingly low alarm rate becomes operationally significant when multiplied across a fleet operating every day of the year.
This illustrates the operational value of pushing nuisance alarms far below the allowable ceiling. It is not an estimate of actual cruise-fleet alarm performance.
A cruise ship is an unusually difficult machine to watch
Thousands of exterior edges
Balconies, open decks, lifeboats, overhangs and changing hull geometry create multiple viewing angles and potential blind zones.
The background is moving
Waves, spray and ship motion create continuous movement immediately beneath the detection zone.
Day and night behave differently
Sun glare, reflections, darkness and deck lighting can challenge optical systems in different ways.
Humans are not the only things falling
Birds, clothing, furniture items, hats and other objects can cross the same basic detection envelope as a person.
Sensors degrade differently
Rain, fog and storms affect radar, thermal imaging, LiDAR and conventional cameras differently, encouraging multi-sensor designs.
The crew has to trust the alarm
High detection probability means little if nuisance alarms train watchkeepers to question the next warning.
The technology is moving from one sensor to sensor fusion
| Technology Layer | What It Sees | Primary Strength | Primary Limitation | Role in MOB Detection |
|---|---|---|---|---|
| Daylight Camera | Visible image and movement | IDENTIFICATION | Darkness, glare and weather | Visual confirmation and event recording |
| Thermal Camera | Infrared contrast | NIGHT OPERATION | Weather and complex thermal backgrounds | Human classification in low light |
| Micro-Radar | Motion, range and trajectory | 3D MOTION | Needs classification support | Identifies objects crossing the hull envelope |
| Computer Vision | Patterns in camera feeds | CLASSIFICATION | Performance depends on training and operating conditions | Separates likely human events from normal activity |
| Sensor Fusion | Multiple independent inputs | VERIFICATION | Greater integration complexity | Reduces dependence on one imperfect sensor |
| Tracking Software | Casualty position after entry | RECOVERY SUPPORT | Sea state can repeatedly obscure the target | Maintains or predicts casualty location |
2017 proved the concept. 2026 changed the certification argument.
90 days aboard an operating cruise ship
Zelim reported that ZOE achieved a 97% detection rate during simulated MOB tests carried out during the ISO certification program aboard Ambassador Cruise Line's Ambition.
The test period ran continuously for 90 days and included winter operating conditions. Lloyd's Register subsequently awarded ISO 21195:2020 certification to the system in April 2026.
Then it encountered a real incident
Ambassador Cruise Line confirmed that ZOE was running aboard Ambition when an actual overboard incident occurred on August 21, 2026. According to the cruise line and technology provider, the event was detected in real time.
The detail that makes the case particularly interesting is that the system had not yet completed full incorporation into the vessel's operating procedures. Technical certification and operational integration were occurring on different timelines.
So why isn't it already on every cruise ship?
| Barrier | Why It Delayed Adoption | What Has Changed | What Still Remains | 2026 Status |
|---|---|---|---|---|
| Detection Accuracy | Early systems struggled to distinguish human falls from normal shipboard and environmental activity. | Modern multi-sensor and AI systems report detection performance around or above ISO requirements. | Performance must remain reliable across different ship designs and conditions. | IMPROVED |
| False Alarms | Birds, waves, reflections and falling objects could generate excessive alerts. | Sensor fusion and improved analytics have sharply reduced nuisance activations in current systems. | Cruise operators still need confidence that fleet-scale alarm rates remain low. | IMPROVED |
| Certification | Operators lacked an internationally accepted performance benchmark and certified commercial products. | ISO 21195 now provides the benchmark, and full third-party certification has entered the cruise market. | More systems and vessel classes must accumulate operational history. | MAJOR SHIFT |
| Regulatory Enforcement | U.S. statutory language includes image capture as well as detection, while implementation has remained incomplete. | Technology availability is now substantially different from 2010. | The Coast Guard still states the specific provision is not currently enforced. | OPEN |
| Retrofit Complexity | A cruise ship requires exterior sensors, cabling, network integration, processing, power, displays and suitable mounting locations. | Operational retrofits have now been demonstrated. | Every ship class still requires its own coverage and integration work. | ONGOING |
| Procedures | Detection is useful only when alarms connect cleanly into actual bridge and emergency-response procedures. | Real-world installations provide operational experience. | Crew training, alarm ownership, verification and MOB response must be integrated. | ONGOING |
| Fleet Economics | Installation and maintenance have to be multiplied across dozens of vessels. | Sensors and analytics are becoming commercially mature. | Contemporary fleetwide turnkey pricing remains largely private. | OPAQUE |
Cost is real, but public pricing is surprisingly old
In its 2016 request for information on cruise-vessel overboard technology, the U.S. Coast Guard said its best available cost information at the time was approximately $300,000 to install a system on an average cruise vessel, plus approximately $40,000 per year for maintenance.
Those numbers are now a decade old and should not be treated as contemporary quotations. Today's systems may include substantially different cameras, processors, software, networking, tracking functions and integration requirements.
Historical installation estimate
Coast Guard best-available figure used during its 2016 technology information request.
Historical maintenance estimate
Annual maintenance figure cited in the same Coast Guard information request.
Current turnkey fleet price
Publicly available sources do not provide a reliable universal 2026 installed price for a large-cruise-ship system.
Installation is more than hanging cameras on the rail
A full installation can require multiple sensor stations around the vessel, suitable fields of view, power, data cabling, marine-grade housings, processing hardware, networking, recording, alarms and an operator interface.
The installation then has to become part of the ship's safety organization. Someone must own the alarm, verify it, communicate it to the bridge and connect the detection to the vessel's MOB response.
MSC demonstrated why this was difficult almost nine years ago
In 2017, MSC Cruises announced an intelligent exterior surveillance system aboard MSC Meraviglia developed with Bosch and Hewlett Packard Enterprise.
Optical and thermal imagery was processed through two independent image-processing systems. MSC specifically identified waves, sunlight, moonlight and birds as causes of false alerts that the technology had to distinguish from genuine overboard events. After more than 25,000 hours of video analysis and software development, MSC reported a 97% accuracy level.
MSC said at the time that similar technology would eventually be developed and deployed across its fleet. The announcement illustrates the central adoption problem: proving a system on one ship and standardizing it across an evolving global fleet are different engineering projects.
Automatic detection also changes what happens after the alarm
The ship first has to prove somebody went overboard
A missing-person report can trigger cabin checks, passenger searches and review of multiple CCTV feeds before investigators identify the time and approximate position of the fall.
During that delay the vessel continues moving and the casualty continues drifting.
The ship begins with a timestamp and position
Automatic detection can provide an immediate event, supporting rapid bridge notification and preserving the location where the casualty entered the water.
Systems that continue tracking the person can extend that advantage into the recovery phase rather than ending their job once the fall is detected.
At 20 knots, a five-minute detection delay is 1.67 nautical miles
A cruise ship making 20 knots travels approximately 617 metres every minute. If an overboard event remains unknown for five minutes, the vessel travels roughly 3.1 kilometres before the bridge even receives the alarm.
Automatic detection does not guarantee recovery. It changes the starting geometry of the rescue.
The technical objection is weakening faster than the regulatory one
Detection technology was not dependable enough
Congressional testimony and Coast Guard work repeatedly identified accuracy and false-alarm performance as reasons automatic MOB technology had not moved to fleetwide deployment.
Certified systems now exist
The industry now has an international performance standard, independent certification, operational cruise-ship installations and at least one publicly confirmed real-time detection during an actual incident.
What happens when detection is scaled across an entire fleet?
Change the assumptions below. The model compares fleet rollout cost, nuisance-alarm workload and the distance a ship travels before an overboard event is recognized.
Screening model only. The default $300,000 installation and $40,000 annual maintenance figures are historical Coast Guard figures cited in its 2016 request for information and are not current vendor quotations. Nuisance-alarm assumptions are user-adjustable and do not represent the performance of a particular system. Distance calculations assume constant straight-line vessel speed until recognition and do not model a subsequent MOB maneuver, current, wind or casualty drift.
Research basis
- United States Code, 46 U.S.C. §3507, passenger vessel security and safety requirements.
- U.S. Coast Guard Cruise Vessel Security and Safety Act guidance and current FAQ concerning provisions not presently being enforced.
- ISO 21195:2020, Ships and marine technology, systems for the detection of persons while going overboard from ships. Standard reconfirmed in 2025.
- U.S. Government Accountability Office, Cruise Vessels: Most Required Security and Safety Measures Have Been Implemented, but Concerns Remain About Crime Reporting.
- U.S. Coast Guard 2016 Federal Register request concerning the status and availability of overboard detection technology for cruise vessels.
- U.S. congressional hearing record on commercial and passenger vessel safety and the cruise industry's experience with MOB detection false alarms.
- MSC Cruises, 2017 announcement and technical description of the MSC Meraviglia intelligent man-overboard video surveillance system.
- Zelim, April 2026 ISO 21195 certification announcement for ZOE following 90-day trials aboard Ambassador Cruise Line's Ambition.
- Lloyd's Register public confirmation of ISO 21195 certification awarded to Zelim.
- Ambassador Cruise Line and Zelim, August 21, 2026 statement confirming real-time detection of an overboard incident aboard Ambition.
- MARSS MOBtronic technical and operational material covering multi-sensor detection, nuisance-alarm reduction, thermal imaging, radar and event processing.
We welcome your feedback, suggestions, corrections, and ideas for enhancements.
Please click here to get in touch