A robot for the one place your hands can't go
Somewhere in a hospital right now, a doctor stands in a lead apron, absorbing radiation for hours, pushing a thin wire through a patient's blood vessels by feel. Yunfei Cao thinks that job belongs to a machine.
His company, Angio8, is building a surgical robot for minimally invasive vascular intervention - the family of procedures where doctors treat blocked or damaged blood vessels by running catheters and guidewires up through the arteries, guided by live X-ray. It is precise, tiring, radiation-heavy work. The robot's pitch is simple: let the human plan the procedure and let the machine execute the millimeter-scale movements, from a console that can sit outside the radiation field.
Angio8 describes what it is building as a next-generation, first-of-its-kind vascular interventional surgical robot platform - a robot that remotely controls the catheters and guidewires a surgeon would otherwise manipulate by hand. That framing is bold. It is also, for a company founded in December 2021 with a team you could seat around a dinner table, enormously ambitious. Cao is the co-founder and CEO steering it.
The interesting thing about Cao's company is not its size. It is its geography. Angio8 runs its engineering and design out of a research center in Boston and its manufacturing, regulatory, and clinical work out of Shenzhen's Nanshan district. Two cities, roughly seven thousand miles apart, stitched into one product. Cao himself is based in Boston, which makes him the human hinge between the two halves.
Four steps between a surgeon's intent and a wire's movement
Robotic intervention is less about sci-fi autonomy and more about turning a doctor's hand movements into safe, repeatable, radiation-free machine motion. Roughly, it looks like this.
Surgeon at console
The doctor controls the procedure from a seat that can sit outside the X-ray field.
Intent to signal
Hand movements become digital commands, filtered for tremor and overshoot.
Robot drives the wire
Motors advance, retract and rotate the catheter and guidewire with sub-millimeter control.
Live feedback
Imaging and force cues loop back so the surgeon stays fully in command.
From supplier to inventor
Cao did not start by building robots. He started by studying the human body as an engineering problem. Between 2008 and 2013 he earned a Bachelor of Engineering in Biomedical Engineering at Stony Brook University on Long Island - the discipline that treats a heart valve and a hydraulic pump as cousins.
Then came the Boston years. Boston is one of the densest medical-device ecosystems on earth, and Cao spent his early career inside it - across roles connected to Vention Medical, SunMed, China Med Device, and Boston Innovation Technology. These are the companies that design, source, and supply the catheters, components, and contract-manufactured pieces that flow into other people's finished devices. It is a useful, unglamorous education. You learn how a medical product actually gets made, certified, and sold - and where the gaps are.
That vantage point matters. A founder who has only ever imagined a device dreams in features. A founder who has spent years supplying the industry knows what a bill of materials costs, what a regulatory submission demands, and how long the road from prototype to operating room really is. When Cao co-founded Angio8, he was not a first-timer walking in cold. He was a supplier who decided to stop handing parts to other builders and build the whole thing himself.
In December 2021 he did exactly that, planting Angio8 in Shenzhen - a city that can turn a hardware idea into a working prototype faster than almost anywhere - while keeping the design brain in Boston. The Chinese name, 通甪科技, sits alongside the English "Angio8," a nod to angiography and to a number that reads as auspicious. The company set out after what it calls the world's first novel vascular interventional surgical robot.
Venture-scale ambition, dinner-table headcount
Angio8 has raised institutional venture money. Its Series A round, closed in May 2023, brought in a reported tens of millions of RMB from a syndicate that included Oriental Fortune Capital (东方富海), Kangyu Capital, and Ruiqian Capital. The capital was pointed where deep-tech medtech capital always goes: accelerating product development and funding the clinical trials that stand between a robot and an approval. Reporting points to continued fundraising activity through 2025.
What makes the number striking is the denominator. This is a small, focused team - the kind of lean deep-tech outfit where every hire is a bet and every engineer wears several hats. Raising serious money to build a surgical robot with a compact team is not a weakness to explain away. It is a statement about how much can be done with focus, a hardware ecosystem at your doorstep, and a founder who already knows the terrain.
Bars illustrate relative momentum, not audited figures. Reported Series A: tens of millions RMB, May 2023.
The boring acronyms behind a robot in your veins
The magic of a surgical robot is invisible. It lives in the parts nobody applauds - the timing, the safety standards, the low-level protocols that make a machine trustworthy enough to touch a heart's plumbing.
IEC 62304
The software lifecycle standard for medical devices. Every line of code has to be born under it.
RTOS
A real-time operating system, because a wire moving through an artery cannot wait for a lag spike.
UART · I2C · RS485
The quiet communication protocols that let sensors, motors and controllers agree, thousands of times a second.
ADC · DAC · PWM
The bridge between the analog body and the digital robot - reading forces, driving motors, holding steady.
BOM Management
Every screw and sensor tracked. In medical hardware, the bill of materials is a legal document, not a spreadsheet.
Vascular Intervention
Treating blood vessels through tiny incisions instead of open surgery. Less trauma, more precision - if the tools cooperate.
A quieter operating room
Strip away the funding rounds and the acronyms and Angio8's goal is almost tender: make surgery safer for the person doing it and steadier for the person receiving it. Move the surgeon out of the radiation. Take the tremor out of the wire. Turn a physically punishing procedure into a controlled, repeatable one.
The road there runs through clinical trials, regulators, and years of patient work - the unglamorous middle distance where most medical-robotics ambitions live or die. Cao is not promising it is finished. He is betting a Boston brain, a Shenzhen body, and a small team can close the gap between a robot that steers a wire in a lab and one that does it in an operating room. That is a strange, specific, worthwhile thing to spend a career on.