A shipowner’s field guide to hull-cleaning robotics

Intro

As a marine coating chemist, my initial reaction to the widening use of robotics (for want of a better word) was annoyance - annoyance from a position of pride, in that chemistry could not solve fouling and needed some draconian million-dollar toothbrush. It just seemed inelegant.

After I got over that stupid thought I saw the opportunities. If we don’t need to worry about the drag that comes from fouling, we can look at the drag that comes from pushing anything through water at speed. That’s the cool thing about underwater cleaning robotics to me - you solve fouling and can then look at optimizing the surface, and maybe achieve a speed gain.

This article is an overview of the different robotics technologies on the market. It’s not exhaustive, so apologies if yours didn’t make it in here.

Overview

If we go back to what we’re trying to achieve with hull performance, it’s zero (or better) speed loss, complete operational flexibility and 100% reliability. At the same time, we should also have excellent HSEQ in terms of harmful content and solvent emissions. If you could solve all that, I don’t think any owner would be upset when you present a seven-figure invoice.

I do need to mention the coating aspect briefly. As a sign of how important coating compatibility with grooming is, in 2025 Lloyd’s Register handed out the industry’s first full antifouling type approval to GIT and Jotun. Those endorsements, from a class society and a coating maker, are noteworthy, and I imagine more will follow. Class approval is a sign of confidence from experts in the field. But let’s not forget that just because LR says something is OK doesn’t mean there aren’t other approaches that are also OK; that’s part of what this article is trying to communicate.

How to categorize the solutions?

It would be tempting to group these robots either by brand or by cleaning mechanism. Neither helps an owner decide. I think it’s more relevant to do it by the operating model: where the robot lives and when it cleans. That determines port logistics, how often you intervene, whether you need a support vessel, and how much the cleaning depends on autonomy that has not been proven at sea. Mechanism still matters, but it is a property of each approach, not the organizing principle.

If we do it that way, the market splits into three families.

In-transit

I think what everyone wants is an autonomous in-transit system that just takes care of everything on a proactive basis. Seems like an ideal scenario to me. Not even a sniff of being delayed because of cleaning - it just happens while underway, when we have all the time in the world.

Here we have tethered systems like Shipshave’s ITCH, which use a deck winch and hydrodynamic forces: no magnets, no autonomy, just the crew deploying and recovering the unit at service speed. Then there are autonomous systems like NakAI, which self-deploys from an onboard dock while the ship is under way, and Nautica, a self-organizing swarm, both chasing the same goal of cleaning with zero crew involvement. Right now, the tethered version is commercial. The autonomous version looks like it’s coming, just not quite here yet.

Onboard, proactive

The device lives on the vessel and cleans early, at the slime stage (FR20-ish), before macrofouling takes hold. It is triggered by a shore operations center using fouling-risk data. Greensea and Jotun chose this approach, with Jotun’s HullSkater sold inside the HSS coating-plus-robot-plus-monitoring bundle. Greensea IQ’s EverClean pursues the same proactive thesis but as a port-based subscription rather than onboard hardware. The argument is that cleaning little and often beats cleaning hard and rarely. Who would want to brush their teeth quarterly?

Port-based

This is the one everyone already knows - the robot comes to the ship in harbor, usually an ROV run off a service vessel or the pier. It’s the largest, most mature family by far, and it splits up by mechanism. High-pressure magnetic crawlers (Fleet Cleaner, VertiDrive) go for raw throughput on steel. Brushless, variable-pressure waterjets (HullWiper, ECOsubsea) skip contact entirely to protect the coating. Cavitation systems (Neptune, CLIIN) fracture fouling at low pressure. And small high-frequency units (Hullbot) clean gently, but often. Whatever the mechanism, they all share the same constraint: you only clean when you’ve got a port call, and you’re competing with everyone else for the slot.

The field at a glance

There is more to it than where the robot lives and when it cleans. Three questions decide how a system fits your operation: does the crew have to operate it, is it autonomous, and is it run from a control center on the other side of the world? None of those is automatically better. Satellite links like Starlink now make remote operation feasible almost anywhere - a real step up from systems that leaned on shore-side 5G - and handing control to a remote center also lets the owner decide where and when to clean. The control and connectivity column below is where each system lands on that.

Twelve manufacturers, grouped by operating model (in-transit, stationary, or port-based):

Manufacturer (product)	Origin & founded	Cleaning technology	Control & connectivity	Weight & size	Maturity
IN-TRANSIT (LIVES ONBOARD)
Shipshave (ITCH)	Norway · 2019	Tethered hydrodynamic drift; soft brush	Crew-operated from deck winch	~50 kg (two-part, crew-portable); dimensions not disclosed	Commercial; multiple fleet deployments
NakAI Robotics (PLECOS)	Israel · ~2020	Cable-free; UV + soft brush	Fully autonomous; no crew after install	Not publicly disclosed	Paid trials underway
STATIONARY (LIVES ONBOARD)
Jotun HullSkater (HSS)	Norway · 1926 (HullSkater 2020)	Magnetic-wheel crawler; cleaning head	Remote ops center (4G / satellite); no crew cleaning role	~200 kg; ~1.6 × 1.0 m	Commercial; LR type approval 2025
PORT-BASED
Fleet Cleaner (now Fleet Robotics)	Netherlands · 2011	Magnetic crawler; high-pressure waterjet	Remote ops center (Delft); operator-run, support vessel	~2.0 × 1.8 × 0.6 m; weight not disclosed	Commercial; 4 service vessels
VertiDrive	Netherlands · 2008	Magnetic crawler; interchangeable heads (waterjet / UHP / abrasive)	Operator, on-site remote control	~62 kg; 0.75 × 0.52 × 0.72 m (V700)	Commercial
CLIIN Robotics	Denmark · 2016	Magnetic crawler; interchangeable brush / cavitation heads	Operator-run, on-site	~89 kg (35 kg hull tool)	Commercial; hull & cargo-hold lines
HullWiper	UAE (dev. Norway) · 2013	Brushless variable-pressure waterjet; debris capture	Operator-run ROV, on-site	~1,275 kg; 3.3 × 1.7 × 0.85 m	Commercial; dozens of ports
ECOsubsea	Norway / UK · 2008	Soft waterjet ROV; ~97% debris capture	Operator-run ROV, on-site	Not publicly disclosed	Commercial; approved for restricted ports
Neptune Robotics	Singapore · 2018 (founded Hong Kong)	Cavitation waterjet; AI-guided	AI-guided, operator-supervised, on-site	~300 kg; dimensions not disclosed	Commercial; ~60 Asian ports
Greensea IQ / Armach (EverClean)	USA · Greensea 2006; Armach 2021	Non-magnetic (suction); soft-brush grooming	Operator-run subscription; autonomous grooming, no crew	<30 kg; ~0.86 m long	Commercial; GIT coating approval 2025
Hullbot	Australia · 2015	Small free-swimming; soft brush	Autonomous, operator-supervised	~10 kg; dimensions not disclosed	Commercial; high-frequency
Nautica Technologies (HYDRA)	Switzerland · 2024	Autonomous swarm; soft brush	Fully autonomous swarm; no crew	Not publicly disclosed	Pilot; $4M seed 2025

What looks best, and why

Here is how the different families play out against the three things owners ask and care about most.

Reliability at sea. Proven port-based ROVs and crew-operated tethered systems are the safe bets today, because a person can recover the unit if conditions change and the operating envelope is well understood. The fully autonomous in-transit concepts are the most exciting and the least proven. Given that it’s a fully mappable substrate, full autonomy is possible and surely coming soon.

Clean capacity (m²/hr). If raw area per hour is the metric, high-pressure magnetic crawlers lead, in the range of roughly 1,000 to 2,000 m²/hr. Brushless waterjet systems can match that on speed, but the ones built for full debris capture trade throughput for containment and run closer to 300 to 600 m²/hr. Cavitation and soft-brush systems are slower still, often 200 to 400 m²/hr, and they buy coating safety with that lost speed. In-transit cleaning does not really belong on the same scale: a tethered unit can do a full mid-size hull in a single voyage leg with zero port time, which can beat a faster robot you have to wait in line for.

Size, weight, access, and cost. Lighter is more flexible. Man-portable units in the 10 to 90 kg range can be carried aboard and deployed by crew, while the high-throughput crawlers need a dedicated support vessel and a remote operations center, which makes them a port-only proposition. On commercial model, you are choosing among three structures: capital purchase, pay-per-visit subscription where the operator carries the equipment risk, and the bundled coating-plus-robot contract that buys integration at the cost of lock-in. None is cheapest in every case. The subscription model lowers the entry barrier; the bundle removes the coating-versus-cleaning warranty argument; owning the hardware makes sense only at fleet scale.

Net read: for most owners today, a proven port-based or tethered in-transit system is the dependable choice, and the right mechanism follows from your hull material and your coating. The proactive and autonomous categories are where the upside is, and they are worth piloting, but they are not yet where the reliability is.

What actually decides the market

The real winners here will be the ones who land high-profile references and deliver a decent result at a sensible cost. Owners adopt on proof, not promises. In a business where a bad clean can cost you a coating warranty or a hull, nobody wants to go first; everybody wants to go second. I do think that integrated solutions where coating and cleaning come together are strong - this can be through Jotun’s approach or a JV type like GIT. For an owner, having one point of contact and liability is beneficial.

The vendor who gets a respected operator to put a real fleet on the line, and then reports real savings, hands every other owner the evidence they were waiting for. That does more than any amount of marketing.

Winning looks like a good result at a sensible cost, not best in class. Fleets do not run on the best available technology; they run on the option that’s reliable and user friendly, with acceptable performance, acceptable risk, and an acceptable price. A system that saves a decent amount of fuel, leaves the coating intact, and does not wreck the maintenance budget will beat a technically superior one that is harder to buy, schedule, or trust.

---

A note on sources: every claim here comes from public manufacturer disclosures, class-society announcements, and trade-press coverage, current as of mid-2026. Throughput figures and capabilities are vendor-reported and should be verified against the most recent technical data when applied to a specific specification.