The CEO of BeeX plans to bring the next AI arms race directly underwater and cut subsea carbon emissions by 95 per cent

In the popular imagination, the artificial intelligence revolution is often pictured as a sleek humanoid robot or a hyper-intelligent chatbot. However, some of the most critical applications of AI are unfolding in the dark, high-pressure environments of the deep sea — a realm that remains the backbone of global infrastructure yet has stayed technologically stagnant for centuries. 

At the helm of this shift is Grace Chia, CEO of BeeX. Her company is redefining how we inspect and maintain the underwater veins of our world: the energy pipelines, power grids, internet cables, and port terminals that keep society functioning.

By deploying Hovering Autonomous Underwater Vehicles (HAUVs) paired with sophisticated AI, BeeX is breaking a century-old monopoly on marine engineering.

Before founding BeeX, Chia was struck by a glaring contradiction. While robots were marketed as tools to simplify life, she observed that in the maritime world, they were actually making everything prohibitively expensive. “Robots were supposed to help make our lives better,” Chia says.

“What I saw was that it made everything super expensive, and yet the underwater inspections still need to be done.” Historically, the companies that owned massive, expensive support boats controlled the market because, as Chia points out, “you cannot do the robotic jobs without the boat.”

Play

This created an immense barrier to entry. Despite the rising need for underwater work due to rising sea levels and ageing infrastructure, the industry remained a closed shop, isolated to a tiny pool of fewer than 50,000 experts globally. “We cannot let this industry continue to stay stagnant,” she asserts. 

Breaking the boat monopoly

BeeX’s solution was to decouple the robot from the massive support vessel. By creating some of the world’s smallest and most intelligent autonomous underwater drones — such as the A.IKANBILIS, named after the Malay word for anchovy — BeeX has effectively democratised the ocean floor. Their technology allows for critical inspections without the tenfold resource strain of traditional methods.

“Our ‘brain’ is more intelligent and tested across more locations compared to ROVs, and our system is a lot more scalable,” Chia explains. This shifts power from those who own heavy, polluting equipment to those who manage intelligent systems, reducing carbon emissions by up to 95 per cent and operational costs by half.

When deploying AI in these high-stakes environments, the temptation is often to automate every variable. Chia, however, chose a path of deliberate restraint. In an industry where a single error can lead to environmental catastrophe or multi-million dollar infrastructure failure, optimisation isn’t enough. “You cannot optimise for trust as every individual with different cultures assesses these systems differently,” she notes. 

Play

This philosophy led BeeX to a contrarian design choice: the human remains the ultimate decision-maker. While their drones possess advanced autonomy, that autonomy can be toggled off at any moment. “The human decides,” Chia insists.

This human-in-the-loop architecture posed a difficult boundary to navigate, requiring significant additional effort to build intuitive interfaces, but Chia was willing to pay the cost to ensure the technology serves as a partner rather than a replacement. 

Transforming the multiples

A common frustration Chia encounters is the public’s stylised view of AI, often driven more by science fiction than by physics. Many assume an underwater drone should operate with the seamless, wireless freedom of a land-based robot. In reality, the physics of water makes data transmission incredibly difficult.

“Society has a very fancy idea of AI, especially those who do not use or see it daily.” This gap, Chia warns, can lead to uninformed decisions by the general public, as technology often advances faster than society can adapt to it. 

Chia seeks to mitigate this by promoting transparency and focusing on defensive measures, particularly given the geopolitical sensitivity of underwater infrastructure. “We build these systems as a defensive mechanism also because we know of the potential impact of these technologies being used by unfriendly geo-political players,” she adds.

As a founder, Chia draws a clear line between intelligence as assistance and as a substitute. To her, “everything is assistance until the day it can substitute”. By using intelligence to “substitute quantity” and focus the team on quality, BeeX can control multiple robots from a single operations hub. This approach not only makes the work more efficient but also safer.

Chia’s vision for the next five years is one in which underwater work is no longer constrained by decades of specialised experience. “Our methodologies can provide access for more people to the world of underwater work, since experience will no longer be a gatekeeper,” she explains.

This shift will eventually replace carbon-heavy ships with swarms of small, efficient drones. While this might change the nature of offshore roles — which were historically lucrative due to the high physical risk — the trade-off, Chia concludes, is a safer, more sustainable global infrastructure.