BrainCo showcases brain-controlled robot at AI conference

BrainCo, a brain-computer interface developer based in Somerville, Massachusetts, this week demonstrated a platform that translates neural signals into robot commands at the 2026 World Artificial Intelligence Conference in Shanghai. A person wearing a lightweight EEG (electroencephalograph) headset thought about grasping a cap, and a robotic arm reached out and completed the action. The entire sequence—from brain signal detection through AI decoding to robot execution—takes under 200 milliseconds.

The technology operates in three sequential steps. First, an EEG headset picks up the wearer's brain signals. Second, AI algorithms decode those signals to identify the user's motor or control intent. Third, that intent is converted into commands for the robot. During the Shanghai demonstration, the mind-controlled robotic arm completed precision tasks including grasping a cup and picking up an apple. BrainCo designed the platform to work with commercially available robots ranging from humanoid machines to robotic arms and legged robots, eliminating the need for proprietary hardware and allowing integration into existing research and development pipelines.

BrainCo frames its technical architecture as "neuro-embodied-AI." The BCI layer decodes intent from brain signals; the AI layer refines that intent by determining how to break down complex intent into actionable steps; and the robot's own systems handle physical execution. Nyx He, partner and senior vice president of BrainCo, stated: "A decade of BCI research has given us the ability to decode what a person intends to do and translate that into machine action. By integrating brain-computer interfaces, AI, and embodied AI, we believe it will define the next chapter of human-machine collaboration."

Alongside the neural control demonstration, BrainCo unveiled its Embodied AI Data Collection Solution. The company identified data scarcity as a persistent technical bottleneck: teaching a robot to perform complex physical tasks such as folding laundry, assembling components, and handling fragile objects requires enormous amounts of high-quality training data. BrainCo's solution leverages proprietary hardware including a dual-arm wheeled data-collection platform and a high-precision glove. Critically, it also captures EEG data from the human operator, recording not only what a person's hands are doing but what their brain is instructing those hands to do. This combines real-robot execution quality with the scalability of human-centric demonstration, allowing continuous, high-volume data collection anchored in real-world tasks. Founded in 2015, BrainCo currently offers three robotic products: the Revo 3 Dexterous Hand (a 21-degree-of-freedom robotic end effector), the Intelligent Bionic Hand (a five-fingered prosthetic), and the Intelligent Bionic Leg (a smart prosthetic knee joint).

BrainCo's demonstration marks a practical convergence of three technologies: brain-computer interfaces, artificial intelligence, and embodied robotics. The company's decade-long focus on BCI research has enabled it to decode motor intent reliably and quickly enough for real-time robot control—a capability Nyx He, the company's partner and senior vice president, frames as defining "the next chapter of human-machine collaboration." The framework BrainCo calls "neuro-embodied-AI" layers intent decoding (via BCI), intent refinement (via AI), and physical execution (via the robot itself) into a pipeline that processes brain signals in under 200 milliseconds.

The introduction of the Embodied AI Data Collection Solution addresses one of robotics' most acute technical bottlenecks: obtaining sufficient high-quality training data for complex, dexterous manipulation tasks. By capturing EEG signals alongside hand and arm movement data—and combining real-world execution with human demonstration and simulation—the system aims to decouple data collection from reliance on large amounts of real-robot interaction. This approach may help reduce the cost and time required to train robots for tasks like assembly, laundry folding, and fragile-object handling. BrainCo's existing product portfolio (Revo 3 Dexterous Hand, Intelligent Bionic Hand, Intelligent Bionic Leg) shows the company has been building both robotics hardware and neural interface expertise in parallel, positioning it to integrate these capabilities into a cohesive platform.

No comments yet. Be the first to share your thoughts!