Hong Bo: Beijing Patient Controls Shanghai Robot via Brain-Computer Interface to Deliver Water
Not a scene from a sci-fi movie: a month ago, a spinal cord injury patient in Beijing's Tongzhou district, sitting in a wheelchair and wearing a domestic NEO brain-computer interface (BCI), controlled a Zhiyuan robot in Shanghai's Zhangjiang, 1000 kilometers away. The patient made the robot pick up a bottle of water, walk forward, and hand it to a visiting young man.
On September 16, at the Shanghai Science and Technology and Journal Frontier Academic Annual Meeting, Hong Bo, a professor at Tsinghua University and chief BCI scientist at Lingang Laboratory, showcased this real footage. He revealed the team's latest progress and the ongoing Avatar Project. Hong Bo said that although the scene is not as cool as in the movie 'Avatar,' the technology chain has been fully connected.
The Avatar Project aims to enable disabled patients to work remotely. Its name comes from the movie 'Avatar,' in which humans transmit brain signals in real time via BCI to control avatars on the planet Pandora. Hong Bo revealed that the original intention was to help motor-disabled patients work and participate in social life without leaving home.
In his envisioned future, these patients can issue commands through a BCI to remotely control humanoid robots. The robots can go to restaurants, exhibition halls, and other places on their behalf to complete service, explanation, and other work. Hong Bo said that although the actions that can be performed now are still simple, he believes that in the future they will unlock more abilities with the help of BCI. He called on more BCI and embodied AI companies to pay attention to this group, noting that even a small monthly income would be a huge help to these patients' families.
Clinical data for the NEO BCI: In March this year, the NEO BCI system developed by NeuroXess (Shanghai) was approved for market launch, becoming the world's first invasive BCI medical device to enter clinical application. Before the product's launch, 32 implant surgeries completed by 11 medical institutions across the country showed that the clinical response rate for grasp function with BCI assistance was 100%.
Hong Bo brought the latest data: a one-year follow-up of the 32 patients showed that 81.5% had an improvement of more than 6 points in the ARAT (Action Research Arm Test) score for single-hand bare-handed movements. This means that one year after implantation, even without BCI assistance, these patients can perform at least two hand movements they could not complete before. Hong Bo said this indicates that a well-designed BCI under scientific principles has the potential to become a method to promote nerve repair.
Hong Bo's imagination for BCI goes beyond helping patients move again. He said that product entry into clinical application does not mean the work is done; for the longer term, countless basic research questions await breakthroughs. He outlined three stages of BCI development to the attendees.
Version 1.0 helps patients go from unable to move to moving. Version 2.0 helps patients' bodies gradually recover. By version 3.0, humans can connect with artificial intelligence through BCI, build human-machine trust, and ultimately move toward human-machine symbiosis.
Hong Bo envisioned that perhaps 10 or 20 years later, instead of everyone having a mobile phone, everyone might have a micro BCI. At that time, people would not even need to type on a computer while listening to a speech. Just as 'atmosphere programming' is popular now, in the future it might become 'atmosphere brain control.' In a further future, humans might lie on Earth and control robot avatars to explore Mars and the unknown universe.
The 1000-kilometer brain-controlled water delivery from Beijing to Shanghai is a demonstration of the Avatar Project landing. It connects BCI and embodied AI. The BCI is responsible for reading intentions, while the robot is responsible for executing actions, with 1000 kilometers of network latency in between. The fact that this chain works shows that the coordination between the two technologies at the communication level has been established. The next question is whether this system can move from demonstration to continuous use, and whether patients' remote work through it can run in real jobs.