News Reporter |
News Editor | Wen Shuqi
In China’s brain-computer interface (BCI) track, motor control has long been the dominant direction. Even Liu Bing, founder of Mingshi Brain-Computer Interface and a Ph.D. from the Chinese Academy of Sciences, initially started his research in motor control. But in the end, they chose a far more innovative path — visual reconstruction.
Founded in October 2024, Mingshi Brain-Computer Interface is a tech company focused on developing medical-grade invasive BCI products. In April 2026, the company announced it had completed a series of angel, angel+, and angel++ funding rounds, raising a cumulative 150 million yuan.
As a company dedicated to invasive BCI technology, Liu Bing understands that “patients only turn to implantable brain-computer interfaces when they’ve truly exhausted every other option.” According to the World Health Organization’s 2019 World Report on Vision, more than 2.2 billion people globally suffer from visual impairment or blindness. China has the largest visually disabled population in the world, with approximately 8.3 million people who are completely blind — about 21% of the global total. For the visually impaired community, the loss of vision is precisely such a hard, almost unsolvable need.
Domestic BCI research started relatively late, and for most companies, following a proven direction is the safer bet. But Liu Bing believes that “the BCI field has reached a juncture where progress can only be driven by innovation.” What BCI entrepreneurship truly needs now, he argues, is “something new that comes from deep thinking and long-term commitment.”
For Mingshi Brain-Computer Interface, that innovation is the direction of visual reconstruction.
So far, Mingshi has completed validation for complex shapes, colors, and grayscale visual reconstruction. Their next milestone is to advance toward implant trials in fully blind patients. If all goes well, they plan to pursue clinical registration in 2028 and bring the product to market by 2030.
“Our ultimate goal is to help those who are completely blind and have no other options reach a level of vision that is medically recognized as sighted,” Liu Bing said.
Visual Reconstruction Is a Blue Ocean Market
In 2016, Elon Musk’s BCI company Neuralink was founded, with motor control as its primary focus. Because China’s BCI research started later, Neuralink — as the industry’s most prominent player — has become a key reference for many BCI startups. But Liu Bing has his own perspective on the feasibility and necessity of motor control.
“There are still some fundamental problems in the motor control field without real solutions,” Liu Bing said. First, regarding long-term effectiveness: signals from motor control implants tend to degrade within about six months, making long-term use impossible for patients. Second, motor control requires actuators that must be compatible with the human body — for instance, patients need robotic arms to grasp objects. But currently, no actuator has been specifically designed for BCI use, so safety, precision, and long-term reliability are all in question. Finally, controlling the actuator itself demands significant attention from the patient — figuring out how to achieve safe, stable, and long-lasting control with minimal cognitive load remains a major challenge.
In contrast, the technology behind visual reconstruction is comparatively more mature, and the partners at Mingshi Brain-Computer Interface all bring prior experience in vision-related fields. “Back then, almost no one in China was doing similar research, so we identified this as a blue ocean market,” Liu Bing said. That’s how they settled on visual reconstruction as their direction.
In 2024, Neuralink publicly unveiled its Blindsight project for the first time, aiming to help completely blind individuals regain basic visual perception by stimulating the visual cortex. This new direction from the industry’s most prominent BCI company aligned remarkably well with the path Mingshi had already chosen. Some of the system parameters Mingshi later published were also strikingly similar to Neuralink’s.
“We were thrilled to see this,” Liu Bing said. “Because we published these things first, and the overlap with Neuralink’s direction proves we’ve been thinking along the right path.”
Today, the visual reconstruction track is fast becoming the new global battleground in BCI competition. According to data from IIM Information, a tech industry research platform, the global visual cortex implant market reached $480 million in 2025, up more than 4.3 times from $90 million in 2020. The China Academy of Information and Communications Technology (CAICT) predicts that China’s BCI market will hit 12 billion yuan by 2030. In 2025 alone, 11 companies and 7 top-tier research institutions worldwide secured over $1.4 billion in funding in the visual cortex implant space.
“Neuralink isn’t just doing motor control either,” Liu Bing noted. Currently, Neuralink’s BCI research is divided into three areas: the first step is “Telepathy” (motor control), the second is “Blindsight” (visual reconstruction), and now the company has proposed a third step called “Deep” — modulating and intervening in brain neurons. “In a sense, we’re essentially benchmarking ourselves against their second step.”
In Medical Devices, Slow Is Fast
Mingshi’s next critical milestone is conducting an implant trial in a fully blind patient, tentatively scheduled between the Chinese New Year period and next summer. After that comes the three-stage timeline of “clinical registration in 2028, market launch in 2030.” This timeline is derived from the clinical registration process, with additional constraints coming from ethical reviews, clinical testing, and regulatory approvals.
Currently, China’s regulatory framework for BCI products is still in the exploratory stage. On March 13 of this year, Boruikang’s implantable BCI hand motor function compensation system received approval from the National Medical Products Administration (NMPA), becoming the world’s first approved invasive BCI Class III medical device.
The approval of Boruikang’s product provides a valuable reference for other BCI companies. But Boruikang’s product is primarily based on motor control. Questions like — which patients are suitable candidates for implantable BCI? What criteria should be met to prove that treatment is genuinely effective? — are details that Mingshi will have to figure out on its own.
On this long and winding road, Mingshi must maintain several delicate balances to reach its ultimate goal.
The first balance is between company survival and long-term development. The cycle from R&D to market launch for BCI products is extremely long, during which the company will inevitably face funding and survival pressures. To address this, the company plans to release some phased products. But phased offerings can’t truly solve all the problems — the ultimate goal still requires long-term research to reach.
The second balance is between scientific thinking and industrialization thinking. First, getting from a research topic to a consumer product that can enter hospitals is no easy feat at any step. “Scientific research often only addresses a very narrow aspect of the problem. Turning that into an engineering prototype is already challenging, and jumping from an engineering prototype to a real product is another massive leap. Then you have to go through medical registration and approval, and finally face the challenge of getting into hospitals,” Liu Bing said.
On top of every link in this chain, there’s yet another layer of difference — the gap between scientist thinking and industrialization thinking. Mingshi’s core team comes from strong research backgrounds, but when it comes to industrialization, “we still need to overcome a lot of inertia in our thinking and in the way we do things.”
In many cases, the market wants products to launch and iterate faster. “But for medical devices, speed means there are things we don’t fully understand, and that’s where problems arise.” Liu Bing can feel the market’s impatience, but he doesn’t subscribe to it. “The market is very eager right now, but I genuinely believe that from a market perspective, we shouldn’t rush. Slow is fast. It’s not true that quickly making something and throwing it out there means it’s necessarily good. Device safety comes first — we need to build things faster, but on a stable foundation.”
Another expectation the market holds for BCI is its potential as a new form of human-computer interaction. In 2026, brain-computer interfaces were written into the government work report and the 15th Five-Year Plan for the first time, officially designated as one of six future industries.
When the conversation shifts from medical applications to broader imagination, Liu Bing maintains the same cautious stance. Regarding BCI’s future consumer-facing applications, “First, we need to clarify whether we’re talking about implantable or non-implantable BCI.”
He explained that for implantable BCI, the first priority will always be medical applications. “Implantable BCI will find it very difficult to become a true consumer product that interacts with everyone in the foreseeable future. Its development path will take longer — after we can confirm that it cures the relevant patients, if long-term safety and effectiveness can be verified, then maybe it can move forward. But for it to enter the daily lives of ordinary people, I think it will take 20 to 30 years, or even longer.”
For non-invasive BCI, Liu Bing pointed out a common misconception. “People often mystify brain-computer interfaces, hoping to use simpler technology to accomplish more complex things. On the technology side, they want it as simple as possible — no trauma whatsoever — but on the functionality side, they demand extremely complex capabilities, like reading minds and doing all sorts of flashy things. But in reality, all of that is extremely difficult. Non-invasive approaches are far less effective, and even invasive ones can’t yet achieve everything people want.” He admitted that non-invasive BCI is still the subject of many attempts, “but I haven’t seen a clearly defined market for it yet.”
That said, Liu Bing doesn’t rule out the possibility of consumer-grade applications. “My personal judgment is that this direction won’t take too long — I believe people will eventually find a good, suitable direction.” But he emphasized that a clear product definition is the prerequisite. “If the definition isn’t clear, I don’t think it’s going to happen anytime soon.”
