Siemens has advanced autonomous manufacturing with a successful test of a new robotic system. The technology giant deployed a humanoid robot, a machine built to mimic human physical structure, designed for industrial logistics at its electronics factory in Erlangen, Germany. Siemens partnered with Humanoid, a London-based robotics startup, and NVIDIA for this deployment.
The trial featured the HMND 01 Alpha, a wheeled humanoid robot, performing autonomous logistics tasks alongside human workers. This operational test moves humanoid robots beyond controlled research laboratories and directly into functional factory roles. The robot executed specific tote-handling assignments during the evaluation. It autonomously picked containers from storage stacks, transported them across the shop floor, and placed them onto conveyors for further processing.
The companies reported the robot met all target performance metrics. The HMND 01 Alpha achieved a throughput of 60 tote movements per hour and maintained continuous operation for more than eight hours. The system also recorded autonomous pick-and-place success rates above 90 percent. Engineers designed the HMND 01 Alpha specifically for complex industrial environments. The robot combines an omnidirectional wheeled base with advanced physical manipulation capabilities.
These features allow the machine to adapt to diverse tasks and operate safely in spaces shared with human employees. The system relies on KinetIQ, Humanoid’s proprietary AI framework, for its core intelligence. It also integrates NVIDIA’s full physical AI stack to process information and learn tasks.
This stack includes Jetson Thor, an edge computing platform for real-time processing, alongside Isaac Sim and Isaac Lab, software tools for virtual simulation and reinforcement learning.
Siemens executives emphasize the critical necessity of deep operational integration for real-world deployment. The company connected the robot directly to the factory’s existing infrastructure using Siemens Xcelerator, its portfolio of industrial software and digital platforms.
This integration incorporated digital twins, exact virtual replicas of physical assets, AI-enabled perception systems, and real-time data exchange protocols. The connection allowed the robot to coordinate workflows with programmable logic controllers, autonomous vehicles, and human operators. The system functioned as a collaborative asset rather than an isolated automation tool.
The successful trial builds upon a proof-of-concept announced earlier this year. The deployment aligns with a shared vision between Siemens and NVIDIA to build fully AI-driven and adaptive manufacturing sites.Executives from the involved companies view this test as a concrete signal of industrial readiness.
Physical AI, the discipline of training intelligent machines to perceive and act in the physical world, now meets the practical demands of modern factory logistics.
Sources: