AI Companies Look to the Ocean as a Place to Put Data Centers

Press handout image courtesy Microsoft

Published Aug 23, 2026 2:34 PM by The Conversation

[By Nir Kshetri]

The artificial intelligence boom is driving unprecedented demand for data centers, raising concerns about their growing energy consumption, water use and carbon footprint. These challenges are making companies look beyond traditional land-based data centers.

Some developers are now exploring the ocean as a new location for AI infrastructure in the hopes that underwater data centers could improve energy use and cooling efficiency while using less fresh water and land area than onshore buildings.

My research focuses on the societal, organizational and environmental implications of emerging technologies, particularly artificial intelligence and the digital infrastructure – including data centers – that supports its development and deployment. I see underwater data centers as a promising new approach for supporting the growth of AI.

But moving servers into the ocean does not make other underlying environmental challenges such as energy consumption and carbon emissions disappear. And it creates new concerns about harm to the marine environment, as well as questions about how these data centers can be regulated – and how companies can maintain and expand them if needed. Whether underwater data centers can become sustainable alternatives to traditional data centers depends on solving these economic, technical and environmental problems.

The rise of ocean-based AI infrastructure

In 2015, Microsoft launched a research project to explore the feasibility, benefits and challenges of underwater data centers. Part of that effort included setting up a waterproof data center on the seafloor near Scotland’s Orkney Islands in 2018. It contained 864 servers and was connected to shore by an underwater cable.

After two years, Microsoft reported that the servers in the underwater data center failed at about one-eighth the rate of servers in comparable land-based data centers. The company is still studying the possible reasons but hypothesizes that in a sealed underwater environment the equipment is less exposed to oxygen, humidity and temperature fluctuations – as well as less jostling from people working to replace broken components.

However, Microsoft ended the project in 2024 and chose not to build more underwater data centers. The company didn’t say why, but others’ analyses suggest the reasons could include regulatory concerns, including the need for environmental permits, as well as a desire for faster upgrades and replacements for the computer equipment inside.

Instead the company has focused on land-based data centers, which can be larger and easier to expand, and also easier to access to repair or replace equipment.

Others have moved ahead, though. China built what may be the world’s first wind-powered underwater data center in Shanghai. The facility launched in June 2025 and began full commercial operations in May 2026.

The US$226 million project uses seawater as a coolant rather than have to refrigerate fresh water, reducing the electricity required to cool the computers. It uses at least 30% less electricity than traditional data centers, and offshore wind turbines reduce reliance on fossil fuels and cut the data center’s carbon emissions.

Japan is testing a different approach: Data centers housed in containers on floating platforms at sea can use seawater for cooling, have unobstructed conditions for solar panels and wind turbines, and reduce demand for land. In 2025, a data center in shipping containers opened on a floating platform near Yokohama. Its power comes from solar panels installed on the same floating platform, with batteries providing energy storage. The test will continue through March 2027.

Singapore is also moving toward commercial-scale floating data centers. In 2026, infrastructure company Keppel began building a four-story floating data center, scheduled to open in 2028. The project will use seawater for cooling, reducing reliance on treated water and improving cooling efficiency. And the fact that it floats means it won’t take up any of Singapore’s limited land availability.

In 2025, Ulsan, South Korea, began planning an underwater data center that could house more than 100,000 servers and use 30% less power than land-based centers by using seawater for cooling.

In Maine, DeepGreen Western Passage has proposed a submersible AI data center in the Bay of Fundy, powered by tidal turbines designed to harness the area’s strong tidal currents.

Land-based data centers could also take advantage of seawater cooling. In Portugal, the SIN01 AI data center in Sines uses seawater from the Atlantic to cool its servers before returning it to the ocean.

The promise of ocean-based data centers

These various approaches offer ways to reduce demand for grid-supplied electricity for powering data centers’ computers and cooling equipment, as well as using less fresh water.

The distance from people’s homes could also be an advantage for data centers in or on the ocean. A Gallup poll in March 2026 found that 70% of Americans oppose building AI data centers in their communities. However, more than half of the world’s population lives within 120 miles of a coast. Underwater could be another way to keep data centers physically close to users for speedy service.

Maintenance, though, is a major challenge. If a computer fails underwater, it cannot be repaired or replaced on site. The entire sealed data center module may need to be brought to the surface, even if just one computer needs work.

Can the ocean sustain AI?

The main environmental concern about ocean-based data centers involves the seawater used for cooling. Discharging warm or hot water can potentially affect oxygen levels, pH and marine life in the surrounding waters.

That heat is already apparent at the few seaborne data centers now operating. HiCloud, the engineering contractor for China’s Hainan underwater data center, has reported a temperature increase of less than 1 degree Celsius (1.8 degrees Fahrenheit) in the seawater near the facility. SIN01 in Sines, Portugal, also returns seawater about 1 C warmer.

Many marine species depend on stable water temperatures for breeding, feeding and migration, raising concerns that heat released by multiple underwater data centers could create localized thermal pollution and alter marine ecosystems. And the ocean is already under pressure. UNESCO, the United Nations agency for international cooperation, including in conservation, estimates that about 60% of marine ecosystems are already degraded or used unsustainably.

The ocean is already warming along with the atmosphere, without additional waste heat from data centers. That additional heat is already threatening coral reef and mangrove ecosystems, seagrasses and other aspects of the marine food web. As that warming continues, ocean waters will be less useful for cooling electronic equipment in some regions.

Underwater data centers could help AI grow while easing some of the pressure on land, energy and water. But the real test is whether the ocean can become AI’s next computing frontier without becoming its next environmental problem.

Nir Kshetri is the Charles A. Hayes Distinguished Professor at the University of North Carolina-Greensboro and a research fellow at Kobe University, Japan.

This article appears courtesy of The Conversation and may be found in its original form here. 

The opinions expressed herein are the author's and not necessarily those of The Maritime Executive.

ClassNK Launches 'Survey Compass,' AI Assistant for Surveys

ClassNK introduces an AI tool to enhance survey operations for shipowners and management companies.

Published Aug 25, 2026 3:18 PM by The Maritime Executive

[By ClassNK]

ClassNK has launched ‘Survey Compass,’ an AI assistant that supports shipowners and ship management companies in their survey-related operations.

Survey Compass is an AI assistant that responds in natural language to questions on classification survey matters faced by shipowners and ship management companies. Users can quickly confirm the information they need together with its supporting sources, thereby improving the efficiency of survey preparation and response work and reducing their workload. This initial release marks the first step toward ClassNK’s vision of a digital service platform for ships in service. ClassNK will continue to expand and enhance Survey Compass together with our customers, incorporating feedback and requests gathered through its practical use.

Survey Compass is available on ‘ClassNK Customer Hub - Ship in Service -’ (hereinafter ‘CCH SiS’), a web portal service for shipowners and ship management companies.

In the maritime industry, regulations, particularly those related to environmental compliance, are becoming increasingly complex, requiring shipowners and ship management companies to manage a growing number of requirements and an ever-expanding volume of information. Against this backdrop, personnel responsible for survey-related matters are often required to identify the applicable requirements within a limited timeframe. To provide prompt and accurate responses to the classification survey-related inquiries received daily from shipowners and ship management companies, ClassNK has developed Survey Compass, an AI assistant specialized in classification survey support.

Survey Compass accepts questions in natural language, directly searches classification survey-related documents, and presents a concise summary together with links to the underlying information sources. Because information related to classification surveys and ship management is closely tied to the safe operation of vessels, a high level of accuracy and reliability is required.

For this reason, Survey Compass adopts a retrieval-based approach that generates answers based on relevant source documents, reducing the risk of AI-generated inaccuracies. Through this approach, it aims to serve as a trustworthy AI assistant that users can rely on with confidence.

By providing links to the source documents used to generate each answer, Survey Compass enables users to review the original references directly and make informed professional judgments. In the first version, the service will start by covering Part B of the Rules for the Survey and Construction of Steel Ships, the Guidance for Undergoing Surveys, and Technical Information.

Based on feedback and usage during the beta phase, ClassNK will continue to enhance Survey Compass by expanding the range of covered documents, including international conventions and flag State requirements, and by linking with vessel-specific data to provide more context-aware support. Through these enhancements, Survey Compass will evolve to better support users in their day-to-day operations. Going forward, ClassNK will continue to expand services through ClassNK Customer Hub to support the digitalization and operational efficiency of shipowners and ship management companies.

Customers using NK-SHIPS can log in to the web service portal on the ClassNK website with their existing account and access Survey Compass via the link to 'ClassNK web service portal.'

The products and services herein described in this press release are not endorsed by The Maritime Executive.

 

Fertility Rate Curbs Japanese Shipbuilding Boom

Courtesy Mitsubishi Heavy Industries

Published Aug 23, 2026 9:22 PM by The Maritime Executive

The Japanese government’s plan to double domestic shipbuilding capacity by 2035 is creating a wave of investment in the sector, attracted by the prospect of sharing a government injection of $6.5 bn in funding and tax incentives. The plan being advanced by the government of Japanese Prime Minister Sanae Takaichi seeks to reverse a decades-long decline in shipbuilding, and forms part of a wider strategy to revive the world’s fourth biggest economy by investing in 17 key security-related areas which in turn will boost Japan’s ability to defend itself. Alongside shipbuilding, the government is also seeking to build stand-alone capability in Artificial Intelligence and semi-conductor design and manufacturing.

According to OECD figures, the Japanese shipbuilding industry held 11% of the global market in 2024, behind China’s 71% and South Korea’s 17% for South Korea. Notwithstanding having slipped in the global rankings, Japan has retained niche market share in the sophisticated segment where vessels must meet demanding compliance or technical requirements, whereas China has gone for a low-cost, high-volume model.

From its heyday in the 1970s, when Japan built about half of the world’s ships, over the last 20 years Japanese yards have suffered a gradual erosion in orders, forcing midsized yards to diversify into other sectors and into ship-repair work. The historical strength of shipbuilding in Japan was built on its need as an isolated global trading island to have a large merchant marine. The Takaichi plan to revive shipbuilding is built on the same premise and on the need to foster sovereign strategic capabilities, the relevance of which has been emphasized by military expansionism in China, but also by the Trump tariff wars and conflict in the Gulf.

Progress with rolling out the plan has been seen in a number of areas.

Namura Shipbuilding is planning to build a dry dock for large vessel construction at its site in Imari Bay on the northwestern coast of Kyushu, north of Nagasaki. The dry dock is to be completed by 2035. Namura in the interim is planning to revive the building of LNG carriers, with Imabari Shipbuilding on Shikoku and Kawasaki Heavy at Kobe planning to do likewise, such that Japan will be producing between three and five LNG carriers per year.

Expansion of merchant shipbuilding is competing for scarce numbers and skills in shipyard labor forces with orders for warships. Mitsubishi Heavy Industries for example have taken three-ship orders for Mogami Class frigates this year, both from the Royal Australian Navy and the Japan Maritime Self-Defense Force (JMSDF). Japan Marine United is busy rolling out new Sakura-class offshore patrol vessels, and there is an active and well-funded program to keep the over 100-strong JMSDF fleet refurbished and replaced as necessary.

Shortage of the necessary skilled workforce is behind this year’s 15% fall in shipbuilding orders, despite the push from Tokyo. Imabari Shipbuilding president Yukito Higaki has said his yards are struggling to meet Japanese replacement orders, let alone having the capacity to take up the rising demand for export orders or having spare capacity to help with the backlog of US Navy repair and refurbishment work. So his company, like others, is intensifying its search for labor-saving technology as well as recruiting skilled workers from overseas.

Even so, it has been estimated that Japanese shipbuilders will need an additional 12,000 workers if they are to meet the target of doubling tonnage delivered. Japan’s dramatically falling birthrate - which hit a record low of 1.14 in 2025 - is shrinking the size of the available workforce every year, and labor shortages are likely to be the biggest impediment to the government’s shipbuilding expansion plans.

 

Assessing hurricane wind damage from above

Image analysis of blue-tarped rooftops underscores the effectiveness of stronger building codes

Kyoto University

Facebook

XLinkedInWeChatBluesky
Email

 

image: 

Color and shape in the drone image on the left are analyzed through region segmentation, and the RGB information from image segments identified as blue tarps is then used to predict the entire damage on the right.

view more 

Credit: KobeU / David Wolf

Kyoto, Japan -- Massive storms like the 2005 Hurricane Katrina, that cause over $100 billion in damage, were once considered once-in-a-lifetime events. Increasingly, however, they are becoming the norm: a direct result of climate change, which has made large storms more frequent, intense, and slower-moving, deepening their impact.

Hurricanes can damage homes through flooding and strong winds, and insurance claims target both. After the 2022 Hurricane Ian, for example, 44 percent of insurance claims were for flooding and 56 percent were for wind damage. Yet most economic research focuses on flooding because elevation data and hydrological models can predict flood extent, whereas identifying the location and severity of wind damage across a wide area has posed a particular challenge.

Unlike flooding, there is nothing like a dam or levee that can shield an entire community from wind, so the responsibility falls almost entirely on individual homeowners. To mitigate wind damage, Florida enacted the 2001 Florida Building Code -- 2001 FBC -- which required new and re-roofed homes to use stronger roof-to-wall connections, thicker roof decking, and even impact-resistant windows and doors in some areas. Many homeowners had to spend thousands of dollars to reinforce a roof or install impact-resistant features, yet whether these preventative measures proved to be worth the average cost of about $4,930 had remained unclear.

This motivated a collaborative team of researchers from Kyoto University and Kobe University to develop a new method of identifying wind damage and assessing the benefit of these measures. The team observed high-resolution drone and satellite imagery taken after a hurricane to locate homes with blue tarps -- the temporary sheeting used to cover damaged roofs -- as a proxy for wind damage.

Applying their research to Hurricane Irma, which struck Florida in September 2017, the team tested how much 2001 FBC reduced wind damage to homes. They used tarp size as a measure of damage severity and corroborated these results with building-permit records, making it possible to map wind damage across hundreds of square miles. With data on more than one million homes across eight southern Florida counties, they then applied a regression discontinuity design to compare homes built just before and after 2001 FBC took effect.

The results revealed that homes built under the new code were about 22 percent less likely to suffer wind damage, and when damage did occur its severity was roughly 27 percent lower. Additionally, damaged homes that were sold before being repaired were both harder to sell and, when they did, sold for around 13.3 percent -- or about $29,400 -- less than undamaged homes.

"This study provides large-scale evidence that stronger building codes are effective for reducing wind risk and adapting to climate change," says co-author Kenji Takeuchi. "I hope this research will provide useful insights for disaster preparedness and the development of effective climate change adaptation policies."

###

The paper "Are Building Codes an Effective Adaptation to Wind Risk? Evidence from Remotely Detected Blue Tarps" appeared on 1 August 2026 in American Economic Journal: Economic Policy, with doi: 10.1257/pol.20240544

About Kyoto University

Kyoto University is one of Japan and Asia's premier research institutions, founded in 1897 and responsible for producing numerous Nobel laureates and winners of other prestigious international prizes. A broad curriculum across the arts and sciences at undergraduate and graduate levels complements several research centers, facilities, and offices around Japan and the world. For more information, please see: http://www.kyoto-u.ac.jp/en

---

Journal

American Economic Journal Economic Policy

DOI

10.1257/pol.20240544 

Method of Research

Data/statistical analysis

Subject of Research

Not applicable

Article Title

Are Building Codes an Effective Adaptation to Wind Risk? Evidence from Remotely Detected Blue Tarps

Article Publication Date

17-Aug-2026

 

Who are the secret gardeners of a parasitic plant?

Kobe University

Facebook

X
Bluesky
Email

 

image: 

Scientists have been wondering who disperses the seeds of Balanophoraceae, a family of parasitic plants growing in the shaded and windless understories of Japanese tropical forests. In his newest study, Kobe University botanist SUETSUGU Kenji provides “the first empirical evidence that hermit crabs can function as both external and internal seed dispersers.” Pictured above is Coenobita brevimanus scraping off the fruits of Balanophora fungosa.

view more 

Credit: SUETSUGU Kenji

Land hermit crabs were caught on camera snacking on the fruits of a parasitic plant. Since the fruits attach to the animals’ bodies and some seeds even survive passage through the digestive tract, the Kobe University study is the first to document land hermit crabs as both external and internal seed dispersers.

In the coastal forests of a tropical Japanese island near Okinawa, there grows an inconspicuous plant that is easy to mistake for a mushroom. It is a non-green parasite feeding on the roots of other plants and produces up to a million tiny seeds, each covered in a minuscule, dry fruit. The seeds, in particular, present a mystery to scientists: How do they get dispersed? “Many people assumed that they simply fall by gravity and are carried away by the wind, but the humid and windless forest floor makes this seem rather implausible,” says Kobe University botanist SUETSUGU Kenji.

His previous studies showed that the fruits of plants in the family Balanophoraceae, with their dull coloration and yeasty scent, are in fact dispersed by invertebrates such as ants, crickets, and cockroaches, with plants whose fruits are firmly attached being visited by insects with stronger mandibles, and those with easily detachable fruits being visited by smaller invertebrates. Balanophora fungosa, too, has dull colors, a yeasty scent and easily detachable fruits, so Suetsugu suspected the dispersers to be ants. But when he covered the plants with nets that let through ants but not larger animals, he found that the plant’s fruits were no longer picked up. So, the Kobe University researcher set up time-lapse cameras to observe who the real disperser was.

The main disperser turned out to be the land hermit crab Coenobita brevimanus, as Suetsugu now reports in the journal Ecology. He found that the crabs scraped off the fruits with their pincers, sometimes removing a plant’s entire fruit ensemble overnight. As a result, some of the fruits ended up attached to the crabs’ bodies, and Suetsugu spotted animals with their cargo even several meters away from the plants. He even found that some of the seeds survived passage through the crabs’ digestive tract, and considering that land hermit crabs can travel up to 100 meters, they are likely effective dispersers of the plant. “I was excited to find this because these animals have usually been regarded as omnivores, scavengers or seed predators rather than potential plant partners,” says Suetsugu. In his newest paper, he writes that “this study provides the first empirical evidence that hermit crabs can function as both external and internal seed dispersers.”

What is more, Suetsugu found that the animals seemed to actively seek out the mature plants, as he didn’t see them around immature plants. This means that the animals were probably attracted by the scent the mature plants give off, and that this scent is a direct advertisement to invertebrates who otherwise feed on almost anything they can find on the ground and get their pincers on.

“When it comes to seed dispersal, large vertebrates have received most of the attention, but small-bodied invertebrates are often overlooked. Nevertheless, they may also help maintain seed movement, especially on islands, in fragmented habitats, and on shaded forest floors,” explains Suetsugu. Every new finding brings him closer to creating a predictive framework for when and why such animals become important dispersers. For land hermit crabs, this is the first observation, but, as he writes in his study, “their abundance in coastal forests and omnivorous feeding habits suggest that this may represent a more general pattern.”

This research was funded by the Japan Science and Technology Agency (grant JPMJPR21D6).

Kobe University is a national university with roots dating back to the Kobe Higher Commercial School founded in 1902. It is now one of Japan’s leading comprehensive research universities with over 16,000 students and over 1,700 faculty in 11 faculties and schools and 14 graduate schools. Combining the social and natural sciences to cultivate leaders with an interdisciplinary perspective, Kobe University creates knowledge and fosters innovation to address society’s challenges.

Balanophora fungosa has dull colors, a yeasty scent and easily detachable fruits, so Kobe University botanist SUETSUGU Kenji suspected the dispersers to be ants. But when he covered the plants with nets that let through ants but not larger animals, he found that the plant’s fruits were no longer picked up.

Credit

SUETSUGU Kenji

Journal

Ecology

DOI

10.1002/ecy.70462 

Method of Research

Experimental study

Subject of Research

Animals

Article Title

Unexpected allies: Land hermit crabs and cockroaches contribute to seed dispersal in a non-photosynthetic plant

Article Publication Date

11-Aug-2026

 

Experienced captains vs. conventional AI: Setting a new course for autonomous ship navigation

AI learns from captains instead of calculations in Japan’s trickiest sea

Osaka Metropolitan University

Facebook

XLinkedInWeChatBlueskyMessageWhatsAppEmail

 

image: 

The AI-generated route (blue) closely follows the course taken by an experienced captain (red) while safely navigating through surrounding ship traffic (gray).

view more 

Credit: Osaka Metropolitan University

Teaching an AI to navigate a ship is far more challenging than teaching it to drive a car. In busy waterways such as Japan's Seto Inland Sea, vessels must safely negotiate dense traffic, narrow channels, hundreds of islands and ever-changing conditions while complying with navigation rules.

To develop an autonomous ship-navigation AI, a research group led by Assistant Professor Takefumi Higaki from the Graduate School of Engineering at Osaka Metropolitan University took a different approach to designing it.

Instead of training their model with objectives and rules, the researchers used the maneuvers performed by the vessel Fukae-Maru, a training vessel from Kobe University. Rather than trying to predict the best action like traditional AIs, the researchers turned to “diffusion AI,” which instead makes decisions based on a range of actions that experienced humans might take and creates a whole trajectory. As the ship operates in the congested Seto Inland Sea, the operators have to make decisions that involve ambiguity and human judgment, sometimes making decisions that are difficult to describe mathematically.

The researchers evaluated their AI against two leading navigation AIs built using conventional machine-learning methods based on imitation-learning. They found that their AI effectively handled situations that confused the other models. It was able to simultaneously handle arbitrary numbers of ships, coastlines, narrow waterways, and speed control in a realistic simulation. When they ran ship encounter tests, it consistently complied with international collision-avoidance regulations, maintaining a safe distance from other ships.

As the model was trained, it began to show unexpected behavior such as performing local navigation customs that were never programmed. A local convention when passing through the Akashi Kaikyo Traffic Route is to keep right within designated traffic lanes. Although this was never explicitly programmed into the AI, it consistently maneuvered the vessel into the correct traffic lane.

“The most distinctive feature of this study is that we did not explicitly balance multiple objectives such as collision avoidance, geographical constraints, navigation efficiency, and compliance with maritime traffic rules; however, the AI achieved them,” Dr. Higaki said. “Our approach enables the AI to autonomously learn sophisticated ship-handling skills directly from real-world operational data rather than having researchers manually define what constitutes correct behavior.”

As more real-world vessel operation data become available and their use continues to expand, autonomous ship navigation systems will become more common, similar to the recent increase in AI-driven cars. The researchers hope that their study will improve navigational safety and help to address the labor shortages that are increasing in the maritime industry, especially in Japan.

The study was published in Ocean Engineering.

###

About OMU

Established in Osaka as one of the largest public universities in Japan, Osaka Metropolitan University is committed to shaping the future of society through the “Convergence of Knowledge” and the promotion of world-class research. For more research news, visit https://www.omu.ac.jp/en/ and follow us on social media: X, Instagram, LinkedIn.

---

Journal

Ocean Engineering

DOI

10.1016/j.oceaneng.2026.125656 

Method of Research

Computational simulation/modeling

Subject of Research

Not applicable

Article Title

Diffusion route planner: Data-driven modeling of human ship navigation that implicitly balances safety, efficiency, and rule compliance under complex geographical constraints