European manufacturer Humanoid has released new information about HMND 01 Alpha Bipedal, providing an update to the humanoid platform the company first introduced earlier this year. The UK-based robotics firm now outlines the robot’s development timeline, technical specifications, and intended deployment path, emphasizing how the bipedal model reached operational readiness shortly after final assembly.
According to Humanoid, Alpha achieved stable walking within 48 hours of assembly. While the company had already announced the platform, the new release details how this was achieved. The team reports that it trained the locomotion system using Nvidia Isaac Sim and Isaac Lab, accumulating 52.5 million seconds of reinforcement-learning data — equivalent to roughly 19 months of real-world training — compressed into two days of simulation. The control policy was then transferred to the physical robot after 3.2 million seconds of training, with limited randomization added for stability under external forces of up to 350 Newtons.
Updated specifications
The new announcement confirms Alpha’s full technical profile. The robot stands 179 centimeters tall and includes 29 degrees of freedom, excluding end-effectors. Its modular design supports either a 12-DoF five-finger hand or a 1-DoF parallel gripper. The head unit integrates six RGB cameras, two depth sensors, and a six-microphone array, while the body includes haptic sensors, joint torque sensing, and force/torque feedback.
Processing is handled by Nvidia Jetson Orin AGX alongside an Intel i9 processor. Power comes from a swappable battery with around three hours of operating time during testing.
Expanded context for use cases
The company reiterates that Alpha is intended for a broad set of tasks across industrial, service, and domestic environments. Founder and CEO Artem Sokolov points to labour shortages in manufacturing and the scale of physical or repetitive work in many sectors. He also notes potential use in home environments, where humanoid systems could assist elderly people or individuals with limited mobility. Humanoid cites global estimates of unpaid domestic and care work to illustrate the scale of demand for support technologies.
Alpha’s capabilities include straight and curved walking, turning, sidestepping, squatting, hopping, running, manipulation, and recovery from pushes. The robot supports interaction through a head display, LEDs, audio output, and a six-microphone array. Its KinetIQ framework — combining vision-language models and action models — provides task reasoning and execution.
Position within Humanoid’s overall roadmap
This update follows the company’s earlier launch of its wheeled Alpha platform, which has already completed initial commercial proof-of-concept projects. With the expanded bipedal specifications, Humanoid outlines a wider application range including warehouse automation, picking, palletizing, and future domestic support tasks.
