The Singapore deeptech cutting the tether on underwater inspection - autonomous drones that survey ships, wind farms and seabeds without divers, ROV pilots or a support vessel overhead.
In 2012, an engineering student at the National University of Singapore kept running into the same wall. The deeper an underwater job went, the heavier and more expensive the machine had to be. Bigger tethers, bigger vessels, bigger crews. Progress in subsea work seemed to mean adding weight, not removing it. Grace Chia started asking a different question: what would underwater inspection look like if you cut the tether entirely?
That question is now a company. BeeX designs, builds and operates a fleet of fully autonomous underwater drones - robots that survey ship hulls, offshore wind farms, floating solar arrays and seabeds on their own, without a diver in the water or a pilot on a cable. Spun out of NUS in 2018 by Chia and co-founder Goh Eng Wei, the company turned more than a decade of university research on Hovering Autonomous Underwater Vehicles into something maritime operators can hire by the job.
Almost everything humans build in the ocean has to be inspected, and then inspected again. Ship hulls foul and corrode. Offshore wind foundations, mooring lines and subsea cables degrade. Ports, terminals and floating solar farms all sit in salt water that never stops working against them. The traditional way to look at any of it is to send down a diver, or a work-class remotely operated vehicle (ROV) trailing a thick cable back up to a crewed support vessel.
That approach is expensive, slow and risky, and it does not scale to the assets being built now. As Chia puts it, the arithmetic of the energy transition is brutal for old-school inspection.
Multiply that by the number of wind farms, solar arrays and cables going into the water each year and you get a maintenance bill that no fleet of divers and tethered ROVs can realistically cover. The bottleneck is not building the infrastructure. It is checking on it, forever.
BeeX's answer is autonomy. Its drones are Hovering Autonomous Underwater Vehicles, or HAUVs - machines that can stop, hold position and maneuver precisely around a structure rather than just cruising in a straight line. The autonomy is what removes the cable. Freed from a tether, the drone does not need a large vessel parked overhead or a pilot steering every movement.
Here is the twist that makes BeeX more than a hardware shop: for most clients, it does not sell the robot at all. It sells the answer the robot brings back. Under an Autonomous Robotics-as-a-Service model, BeeX owns and operates the fleet and delivers the inspection data - 2D sonar mosaics, 3D point-cloud visualisations, corrosion readings - as a service. Clients skip the capital outlay of buying vehicles and crewing vessels.
That choice ties directly to the company's cost advantage. Because BeeX designs, manufactures and operates the whole stack itself, it controls every link in the chain from sensor to report.
The intelligence behind the fleet is a software suite called SAMBAL, which handles autonomous navigation, sensor fusion and the job of turning raw acoustic returns into something a human asset manager can read. The hardware line leans local and maritime in its naming: BETTA, the company's most advanced hovering drone, named after the fighting fish; and A.IKANBILIS, a nod to the tiny anchovies that flavour half of Southeast Asian cooking - small, nimble, everywhere.
The flagship HAUV for close-up inspection and intervention. Booked out through the end of 2025 before it shipped.
Autonomous subsea survey and inspection drone, demonstrated at Oceanology International 2026.
The brain: navigation, sensor fusion and 3D point-cloud data delivery that lets the drones run without a pilot.
Robotics-as-a-Service. BeeX operates the fleet and delivers inspection outcomes, no client capex required.
BeeX works across three markets it groups as defense, offshore energy and coastal infrastructure. On the defense side that means intelligence, surveillance and reconnaissance, minehunting and seabed intelligence. In energy it is asset-integrity inspection for wind, solar and oil and gas. Along the coast it is ports, terminals, decks and construction. The client list spans commercial and government users, with deployments in Europe and Asia.
The pull is strongest where the numbers get large. BeeX has framed its opportunity around markets that traditional methods struggle to service at scale: offshore wind inspection, floating solar, and the ever-present demand for ship-hull surveys. The BETTA drone selling out through 2025 before delivery is the clearest signal that demand is real, not theoretical.
There is a quieter reason the service model matters to these buyers. An offshore operator does not want to become a robotics company. It wants a report that says whether a mooring line is fraying or a turbine monopile is corroding, delivered on a schedule it can plan around. By keeping ownership of the vehicles and the operators, BeeX absorbs the messy parts - maintenance, upgrades, the learning curve of running autonomy in cold, dark, current-swept water - and hands back only the finished picture. For a client, that turns an unpredictable capital project into a line item.
Underwater autonomy is genuinely hard in ways that surface robotics is not. There is no GPS below the waterline, so a drone has to work out where it is from sonar, inertial sensors and whatever landmarks the seabed offers. Visibility can drop to nothing. Currents shove the vehicle off course. Every one of those problems is a place where a tethered ROV would simply lean on its human pilot - and where an autonomous drone has to decide for itself.
This is where BeeX's decade of accumulated operating data becomes a moat rather than a line on a slide. The company has run its vehicles across different geographies, seasons and weather, and each mission feeds the autonomy stack more examples of what the real ocean throws at a robot. To sharpen navigation further, it integrated Water Linked's Sonar 3D-15, tightening the drones' sense of the structures around them. The result is a system that gets more reliable the more it works, which is exactly the flywheel a hardware-plus-software company wants.
BeeX competes on two fronts at once. Against incumbent ROV survey contractors, its advantage is autonomy and the service model - no tether, no dedicated vessel, and results priced as data rather than day-rates. Against other autonomous vehicle builders, its edge is vertical integration: owning hardware, autonomy software and operations end to end, which is what keeps its cost of data below rivals who assemble their capabilities from parts.
The company's history helps explain the confidence. The core team traces back to an NUS student robotics club that won "Best New Entry" at RoboSub in the United States in 2013. Roughly 60% of the team has worked together for more than a decade. That is a lot of accumulated muscle memory pointed at one narrow, deep problem.
BeeX raised roughly US$2 million in seed funding in 2023, co-led by Earth Venture Capital and ShipsFocus Ventures. In June 2026 it closed a US$7.7 million Series A led by Monk's Hill Ventures, with Seeds Capital, NUS Technology Holdings, ShipsFocus and Octave Capital joining. The round funds a global push and a bigger fleet. Around the same time the company unveiled BETTA at a Singapore launch and demonstrated A.IKANBILIS at Oceanology International.
The through-line from a 2012 question to a Series A is unusually clean. Chia did not set out to build a better ROV. She set out to find out what a robot could do once the cable was gone. The market, it turns out, has a great deal underwater that needs checking - and a shrinking appetite for sending people and vessels down to do it.