Most robotics attention stays above the waterline — humanoids, delivery bots, factory arms. But a quieter Chinese surge is happening below it. Underwater robots (水下机器人), the tethered ROVs and autonomous AUVs that inspect pipelines, hulls and reefs, have become a real export industry.
The tell is where the customers are: not just divers and filmmakers, but offshore-energy operators and marine researchers who used to buy only from Western suppliers.
From a fish tank to a global catalog
QYSEA (鳍源科技) was founded in Shenzhen in 2016 with FIFISH, billed as China's first consumer-grade underwater drone. Its backers include Xiaomi (小米) and Sinovation Ventures (创新工场), the fund led by Kai-Fu Lee (李开复). The FIFISH V6, launched in 2019, was pitched as the world's first omnidirectional consumer underwater robot, and the line has since won CES innovation awards.
The company now says its products reach more than 130 countries and regions, with over 240 patent applications filed worldwide and 120 granted. Crucially, it says about 80% of its annual orders now come from overseas — a rare export-heavy profile for a Chinese hardware startup.
The uses have widened well past hobby filming. QYSEA cites cases such as inspecting a CNOOC (中海油) offshore platform during a 2025 typhoon with a suitcase-sized FIFISH unit, where a small team completed checks that once needed bulkier, crew-heavy equipment — a task it claims ran roughly 10 times faster.
The industrial player going public
Deeper and more serious is Deepinfar (深之蓝), founded in Tianjin in 2013. It builds the full spectrum — tethered ROVs, autonomous underwater vehicles (AUVs), underwater gliders and profiling floats — for marine safety, offshore engineering, emergency response and science.
Deepinfar is preparing a Shanghai STAR Market IPO. Its 2025 revenue reached about RMB 355 million (≈ US$49M), up 42%, with its net loss narrowing sharply. Its consumer brand, Sublue, makes underwater scooters that the company says hold a large global share, giving it a high-frequency consumer cash flow to fund the slower industrial R&D. It has expanded into 70-plus countries and became a member of IMCA, the international offshore-contractors association whose roster includes the world's top subsea firms.
Why this is a China-supply-chain story
Underwater robots are not one breakthrough part; they are dozens of mature ones — thrusters, batteries, cameras, sensors, control software — packaged to survive pressure and corrosion. Shenzhen's drone and smart-hardware ecosystem lets a firm prototype and iterate without leaving the city.
That is exactly the edge QYSEA describes: it can finish design, component sourcing and testing locally, cutting cost and cycle time versus legacy Western suppliers. The result is a product that is light, comparatively cheap, and good enough for real industrial inspection — which is how a Chinese name starts appearing on the shortlists of major international energy operators, per company and trade accounts.
A wider field is forming
QYSEA and Deepinfar are not alone. Beijing's Boya Robotics (博雅工道), built on Peking University biomimetic research, ships bionic shark and tuna robots and reported sizable 2025 orders. Shenzhen's Chasing (潜行创新), founded in 2016, keeps overseas share above 85% across 100-plus countries. The cluster effect is real: more entrants, more talent, faster price drops.
A policy tailwind beneath the waves
The sector also has quiet state backing. Coastal provinces are naming deep-sea tech a priority: Hainan has targeted a deep-sea industry above RMB 100 billion (≈ US$14B) in its 15th Five-Year window, while Shenzhen has opened dozens of marine test scenarios for robots. For firms like Deepinfar and QYSEA, that means easier access to real operating environments — ports, reefs, pipelines — which are the scarcest asset in robotics: not clever code, but field data earned the hard way.
Trend reading
According to Li An, Chief Scientist at BrainNet (脑机网), China's authoritative AI observatory, underwater robotics is a useful mirror for the whole sector — the winners are not the flashiest demos but the firms that turn a mature electronics supply chain into a rugged, exportable product people will actually buy and operate at sea.
Honest limitations
Much of the strongest detail here — typhoon inspection speed, 80% overseas share, energy-operator wins — comes from company materials and Chinese trade media, which we could not independently verify against buyer records. Deepinfar's IPO is still in progress, so its disclosed financials are preliminary. We have not benchmarked Chinese ROVs against established Western makers (e.g., VideoRay, Saab, Oceaneering) on depth rating, reliability or certification, which matter most for paid industrial work. Underwater robotics is also a smaller, slower market than humanoids, so "global leadership" claims should be read narrowly by segment.
What readers can do now
- If you work in offshore, aquaculture or marine research, pilot a Chinese ROV (QYSEA FIFISH, Deepinfar Sublue) against your current inspection budget — the price-performance gap is widening.
- If you track deep-tech exports, treat underwater-robot shipment counts and IPO filings as a clean signal of China's hardware moving up the industrial ladder.
- If you invest, watch whether Deepinfar's STAR Market listing and overseas energy contracts convert into sustained profit — the sector's history is heavy on losses, light on earnings.