Recently, a team from the Chinese University of Hong Kong completed autonomous surgery on live pigs using a domestic robot, marking the world’s first instance of such a procedure. This event has indeed occurred. They integrated AI-driven automation technology into a domestic laparoscopic surgical robot, successfully performing tasks such as tissue retraction, gauze grabbing, and vessel clamping without any manual operation from doctors. This achievement was also published in the journal Science Robotics.
Surgical robots have long been referred to as the “crown jewel” of the robotics industry, assisting doctors in millions of minimally invasive surgeries. However, previous robots required doctors to operate from a control console, using both hands and feet. This time, the Hong Kong team enabled the robot to work autonomously, using AI to “see” the surgical scene and complete tasks automatically, which is indeed a significant breakthrough.
The team is led by Assistant Professor Dou Qi from the Department of Computer Science and Engineering at the Chinese University of Hong Kong, bringing together talents from various fields including machine vision, reinforcement learning, engineering, and medicine. They developed an AI-driven embodied intelligence framework that integrates visual models, reinforcement learning, and servo control technology. In simpler terms, they equipped the robot with “smart eyes” that can analyze the surgical environment in real-time using only the video images from the endoscope, without the need for additional sensors. Compared to previous complex solutions that required extra sensors, this purely visual technology is more straightforward and easier to use clinically.
There are two key innovations in their work: first, practicality, as they can accurately understand the scene using binocular video; second, flexibility, as they employ data-driven interactive learning that can adapt to different surgical tasks rather than being limited to fixed actions. They integrated this technology into the Conmed laparoscopic surgical robot and validated it on live pigs at the Medical Robotics Innovation Technology Center of the Chinese University of Hong Kong, where the robot autonomously completed multiple operations with precision and safety. This is the world’s first live experiment of multifunctional surgical automation, paving the way from concept to clinical application.
Dou Qi returned to China in 2020 after completing his postdoctoral research at Imperial College London and established a laboratory at the Chinese University of Hong Kong. At that time, the fields of AI and surgical robots were still operating independently, with little intersection. They started almost from scratch, facing the biggest hurdle of not having a ready-made simulator to adapt AI technology. Therefore, the team developed a digital twin platform called SurRol, which first trains the technology in a virtual environment before transferring it to the real robot. This platform is now open-source for global researchers to use. In 2021, they collaborated with Conmed for hardware and software validation, and later the medical school team joined, leading to the live pig experiments.
From nothing to something, the team gradually gathered talents from various fields, breaking down disciplinary barriers. Zhao Weiren, Dean of the Medical School at the Chinese University of Hong Kong, stated that this interdisciplinary achievement demonstrates the potential of AI in surgical procedures and accelerates the transition from laboratory to clinical practice.
Consider the previous laparoscopic surgeries, where doctors had to control three robotic arms and switch controls with their feet, making it feel like playing a video game. Now, AI can automatically handle auxiliary tasks such as retraction and hemostasis, allowing doctors to focus on core operations. Of course, AI cannot yet independently complete an entire surgery, and Dou Qi has acknowledged that this is unrealistic. However, their goal is to make the robot the “third hand” for doctors, similar to assisted driving in cars, reducing the burden and increasing efficiency. Currently, they are conducting multiple rounds of animal experiments, steadily progressing towards clinical applications. In the future, they also hope to enable the robot to learn different doctors’ operational habits, optimizing human-robot collaboration and truly becoming an intelligent partner.
In our country, the development of surgical robots started relatively late, but in recent years, we have caught up quickly. With the support of AI, we have directly reached the forefront. Consider this: surgery is a matter that concerns everyone. A friend of mine had a relative who underwent lung surgery two years ago using an imported robot. The doctor said it was minimally invasive, with quick recovery and good results. However, the doctor had to monitor the entire process, and the costs were not low. Now that domestic robots can operate autonomously, costs are expected to decrease, making them accessible to ordinary people.
How does China’s progress differ from that of other countries? The global surgical robot market is highly competitive. In July last year, Johns Hopkins University reported in Science Robotics that their AI robot could autonomously complete some steps of cholecystectomy, marking a breakthrough. In June 2025, IEEE Spectrum reported that they developed the SRT-H hierarchical surgical robot transformer, which can adapt in real-time to anatomical features, demonstrating greater autonomy. While impressive, our solution excels in multitasking flexibility, capable of handling various surgical scenarios. As for other countries, systems like Intuitive Surgical’s da Vinci remain the global benchmark, but they primarily rely on doctor control, with a lower degree of automation. In June 2025, MicroPort MedBot’s Toumai robot won an industry breakthrough award, indicating that Chinese companies are gaining recognition globally.
This technology shines not only in the operating room. Consider large hospitals in cities with complete equipment and experienced doctors, but what about remote areas? Medical resources are scarce. If AI surgical robots can be combined with remote control, allowing doctors in cities to guide robots working locally, patients won’t have to travel long distances to major hospitals. This can save time, reduce costs, and even save lives. Looking globally, with an aging population and increasing surgical demands, but not enough doctors, AI can fill this gap.
Our team is progressing steadily, integrating medicine and engineering, and fostering collaboration between industry, academia, and research. In the future, AI can become even smarter, learning different doctors’ operational habits, making the operating room feel like it has a tacit teammate. This reminds me of small changes in life, like how smartphone navigation is becoming more accurate and home appliances are getting smarter; the operating room is no different, as technology makes life more secure. Who wouldn’t want that?
——————————If you find this interesting, please share.