Mind-Controlled Robotics: BrainCo Unveils Neuro-Embodied AI Platform at WAIC

BrainCo, a leading Chinese brain-computer interface (BCI) company, has unveiled the world's first integrated brain-to-robot AI platform at the World Artificial Intelligence Conference in Shanghai, demonstrating a technology that allows humans to control robots using nothing but neural signals.

The platform, which the company claims is the first of its kind, uses a lightweight EEG headset to capture brain signals, AI algorithms to decode motor intent, and a control interface that translates those intentions into robot commands in under 200 milliseconds.

In a live demonstration, a user wearing the headset directed a robotic arm to grasp a cup and pick up an apple without any physical movement, voice command, or button press.

BrainCo describes the underlying framework as "Neuro-Embodied-AI," where BCI decodes intent, the AI layer refines that intent by breaking down complex actions into actionable steps, and the robot's own systems handle physical execution.

Unlike invasive BCI solutions such as Elon Musk's Neuralink, which requires surgical implants, BrainCo focuses on non-invasive wearables that are safer, more accessible, and designed for rapid deployment in real-world scenarios.

"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." – Nyx He, Partner and Senior Vice President of BrainCo

Beyond the brain-control platform, BrainCo also debuted an Embodied AI Data Collection Solution designed to address one of robotics' most persistent challenges: the shortage of high-quality, real-world training data needed to teach robots complex, dexterous tasks such as folding laundry or handling fragile objects.

The solution uses a dual-arm wheeled data collection platform and a high-precision data collection glove that records not just what a person's hands are doing, but what their brain is telling those hands to do.

The integrated platform approach allows researchers without prior BCI backgrounds to complete the entire process from wearing the device to controlling a robot within 10 minutes. Key features include:

• Compatible with multiple robot types – Supports humanoid robots, robotic arms, and quadruped robotic dogs, including Unitree G1 Edu, Realman 6-DOF arm, and Deep Robotics Lite 3
• Two classic BCI paradigms – Native support for motor imagery (MI) and steady-state visual evoked potential (SSVEP) with built-in decoding algorithms
• Graphical interface – Visualized workflows from impedance detection to online reasoning, eliminating the need to write low-level code
• EEG hardware flexibility – Compatible with both wet and dry electrodes, supporting multiple sampling rates from 250Hz to 1000Hz
• Open interfaces – Provides SDKs for data access, algorithm integration, and robot task registration for researchers and developers

As the brain-controlled robotics ecosystem continues to evolve, BrainCo's platform represents a shift from isolated demonstrations to reusable, scalable development tools for thought-controlled machines.