Not Being a Silicon Valley Follower: A Few Young PhD Students Bet on the Integrated Brain of Bipedal Humanoids
A startup team of several PhD students focuses on the integrated brain of humanoid robots, achieving fully autonomous go-kart driving with continuous cornering, and exploring new paths for embodied AI.
The team consists of several PhD students and focuses on the core control system of humanoid robots, known as the "integrated brain." By achieving fully autonomous go-kart driving with continuous cornering, they demonstrated perception and decision-making capabilities in dynamic environments, marking a significant advancement in the field of embodied AI.
Humanoid robots are considered one of the ultimate forms of artificial intelligence, with the core challenge lying in enabling robots to act autonomously in complex physical worlds. The team's choice to bet on an integrated brain means attempting to process perception, planning, and control through a unified AI model, rather than traditional modular stacking. This aligns with the current trend of applying end-to-end large models in the robotics field.
The demonstration of fully autonomous go-kart driving with continuous cornering verified the stability and real-time performance of the algorithm in high-speed motion scenarios. If matured, this capability could be transferred to scenarios requiring complex motion control, such as logistics handling and home services, providing technical support for the commercialization of humanoid robots.
The domestic team emphasizes not being a follower of Silicon Valley, reflecting local innovation willingness in the embodied AI track. As large model capabilities improve, robots are evolving from single-function to general-purpose agents. This kind of technical exploration focusing on underlying control and brain integration will accelerate the process of artificial intelligence moving from the virtual world to the physical world.
This page provides an editorial summary based on publicly available information. It is not a republished article. Use the source link below for the original report.