Humanoid Robot
A robot with a human-like morphology, typically equipped with bipedal locomotion, dual-arm manipulation, and perception systems, capable of performing complex tasks in human environments. It represents one of the frontier directions in contemporary robotics research.
Core Facts
In robot competitions, the controllers held by staff members are not used to directly operate every movement of the robot; instead, they are used to switch between preset task modes and guide the travel route, while the robot autonomously executes the movements.
Sim-to-Real (simulation-to-reality transfer) has become the mainstream industry approach for humanoid robot motion control, with researchers using GPUs to run thousands of simulation instances in parallel in physics simulators to train reinforcement learning agents.
Between 2023 and 2025, the global humanoid robot sector attracted more than 10 billion USD in investment.
Timeline (last 90 days)
Retargeting refers to mapping the motion trajectory of one motion subject onto another subject with different body proportions and joint structures, here transferring from human motion to the joint space of a humanoid robot; because the number of degrees of freedom, link lengths, and joint limits are inconsistent, directly copying joint angles often causes the robot to self-intersect or lose balance.
A manipulator arm with force sensing still costs tens of thousands of dollars, keeping humanoid robots out of most workplaces.
Robotics and Autonomous Systems is an interdisciplinary field involving mechanical engineering, electronics, control, AI, and software engineering.
The action space of a humanoid robot is large and highly coupled, and the legs, arms, and fingers must move in coordination while maintaining balance, making joint-level action learning extremely difficult.
将运动规划与执行跟踪耦合在一起（每帧同时重算动作与平衡）会导致计算量随自由度数量指数级膨胀，人形机器人通常有二三十个以上自由度，极易超出实时计算预算。
High-performance actuators, durable materials, and long-lasting batteries are the sources of humanoid robots' high costs, and the technical cost gap between factory robots and household robots is orders of magnitude.
Assessing progress in the robotics field should focus more on hard metrics such as mass-production cost, mean time between failures (MTBF), and actual deployment scale, rather than demonstration videos.
The AI methods driving the rapid progress of large language models cannot simply be replicated for robots, and emergent capability leaps will not happen automatically.
MIT Technology Review co-published an analysis with the nonprofit research foundation Aventine, pointing out that AI breakthroughs in robotics will not genuinely change the daily lives of ordinary people in the short term.
Current humanoid robots commonly face bottlenecks of short battery life (most lasting 1–2 hours) and long charging times, which constrains their use in industrial scenarios requiring continuous long-shift operation.
40 more timeline events
All Facts (20)
In robot competitions, the controllers held by staff members are not used to directly operate every movement of the robot; instead, they are used to switch between preset task modes and guide the travel route, while the robot autonomously executes the movements.
75%VerifiedSim-to-Real (simulation-to-reality transfer) has become the mainstream industry approach for humanoid robot motion control, with researchers using GPUs to run thousands of simulation instances in parallel in physics simulators to train reinforcement learning agents.
75%VerifiedBetween 2023 and 2025, the global humanoid robot sector attracted more than 10 billion USD in investment.
65%VerifiedThe real-time perception-decision-execution closed loop in high-speed motion remains one of the biggest bottlenecks for humanoid robots transitioning from laboratories to real-world deployment.
65%VerifiedImpedance control is suited to robots with controllable torque (such as direct-drive joint robots), while admittance control is suited to industrial robots that are position-controlled but highly rigid.
65%Unverified当前人形机器人主要应用于仓储物流等相对标准化的场景
75%UnverifiedMIT Technology Review co-published an analysis with the nonprofit research foundation Aventine, pointing out that AI breakthroughs in robotics will not genuinely change the daily lives of ordinary people in the short term.
60%UnverifiedTorque density is a core metric for evaluating the performance of humanoid robot joint actuators, with mainstream solutions including quasi-direct-drive motors and harmonic reducer solutions.
60%UnverifiedThe mainstream solutions for humanoid robot joint actuators include quasi-direct-drive motors and harmonic reducer solutions; the former has a low gear ratio with high force control transparency, while the latter has a high transmission ratio but relatively poor back-drivability.
60%UnverifiedHumanoid robot competitive events typically cover tasks such as walking, balance, obstacle avoidance, grasping, and even more complex collaborative tasks.
60%UnverifiedA typical full-size humanoid robot has 30 to 50 degrees of freedom.
60%UnverifiedRetargeting refers to mapping the motion trajectory of one motion subject onto another subject with different body proportions and joint structures, here transferring from human motion to the joint space of a humanoid robot; because the number of degrees of freedom, link lengths, and joint limits are inconsistent, directly copying joint angles often causes the robot to self-intersect or lose balance.
50%UnverifiedA manipulator arm with force sensing still costs tens of thousands of dollars, keeping humanoid robots out of most workplaces.
50%UnverifiedRobotics and Autonomous Systems is an interdisciplinary field involving mechanical engineering, electronics, control, AI, and software engineering.
50%UnverifiedThe action space of a humanoid robot is large and highly coupled, and the legs, arms, and fingers must move in coordination while maintaining balance, making joint-level action learning extremely difficult.
50%Unverified将运动规划与执行跟踪耦合在一起（每帧同时重算动作与平衡）会导致计算量随自由度数量指数级膨胀，人形机器人通常有二三十个以上自由度，极易超出实时计算预算。
50%UnverifiedThe AI methods driving the rapid progress of large language models cannot simply be replicated for robots, and emergent capability leaps will not happen automatically.
50%UnverifiedHigh-performance actuators, durable materials, and long-lasting batteries are the sources of humanoid robots' high costs, and the technical cost gap between factory robots and household robots is orders of magnitude.
50%UnverifiedAssessing progress in the robotics field should focus more on hard metrics such as mass-production cost, mean time between failures (MTBF), and actual deployment scale, rather than demonstration videos.
50%UnverifiedCarefully choreographed robot demonstration videos typically showcase best performance under ideal conditions, which is vastly different from reliable deployment in real-world environments.
50%