Siemens and Humanoid has introduced what it says is “a landmark milestone” within the journey to deliver bodily AI from imaginative and prescient to industrial actuality.
Humanoid’s HMND 01 wheeled Alpha humanoid robotic, constructed utilizing the Nvidia bodily AI stack, has been efficiently examined in operations at Siemens’ electronics manufacturing facility in Erlangen, Germany, performing autonomous logistics duties. (See video under.)
This builds on the Siemens and Nvidia strategic partnership, introduced at CES, to construct the world’s first absolutely AI-driven, adaptive manufacturing websites.
The daybreak of bodily AI in manufacturing
Bodily AI – the self-discipline of coaching clever machines to understand, cause and act within the bodily world – is poised to rework how items are made.
Bridging the hole between AI analysis and the calls for of an actual manufacturing facility requires a high-performing ecosystem: world-class AI compute and simulation, a confirmed robotics platform, and the deep industrial automation infrastructure to tie all of it collectively.
The HMND 01 Alpha robotic was deployed in Siemens’ logistics operations, the place it autonomously executed tote-handling duties – choosing, transporting and inserting containers for human operators.
All goal efficiency metrics had been met, together with a throughput of 60 tote strikes per hour, uptime exceeding 8 hours, and autonomous pick-and-place success charges above 90 %.
Constructing the commercial spine with Siemens Xcelerator
A humanoid robotic’s true worth is in turning into a totally built-in, collaborative asset on the store flooring. Meaning real-time information alternate with manufacturing techniques and different autonomous guided automobiles, synchronized workflows with different equipment and human operators, and adaptive conduct that responds dynamically to altering circumstances.
With out this deep integration, even probably the most refined robotic stays an remoted characteristic.
Siemens gives this vital layer by its Siemens Xcelerator portfolio, from a complete digital twin to AI-enabled notion, to built-in management and PLC-robot interfaces, together with fleet administration, industrial communication networks and high-performance drives.
Collectively, these applied sciences kind the digital spine and automation infrastructure that assist to make sure humanoid robots function effectively and in live performance with the broader manufacturing facility atmosphere. The result is a factory-grade mannequin for deploying humanoids in any industrial setting.
Accelerating intelligence with Nvidia libraries, frameworks and AI infrastructure
Humanoid has built-in Nvidia’s full bodily AI stack into the HMND 01 platform, together with Nvidia Jetson Thor for edge compute, Nvidia Isaac Sim for simulation and Nvidia Isaac Lab for reinforcement studying and coverage coaching.
The result’s a dramatic compression of growth timelines. Simulation-first {hardware} design has additionally enabled the workforce to optimize actuator choice, joint energy and mass distribution just about, chopping prototype growth from a typical 18-24 months to only 7 months.
Deepu Talla, vp of Robotics and Edge AI at Nvidia, says: “Factories of the longer term demand robots that may understand, cause, and adapt autonomously alongside human employees, tackling the labor shortages and operational complexity that conventional automation struggled to deal with.
“With Siemens offering the commercial integration spine and Humanoid deploying Nvidia’s full bodily AI stack – from simulation-first coaching to real-time edge inference – this deployment paves the best way for humanoid robots assembly actual manufacturing targets on a stay manufacturing facility flooring.”
Humanoid: Constructing factory-grade humanoids
Humanoid, a UK-based AI and robotics firm, developed the HMND 01 Alpha, a humanoid robotic purpose-built for industrial environments.
Combining an omnidirectional wheeled mobility platform with superior manipulation capabilities, powered by KinetIQ, a proprietary AI framework, the HMND 01 is engineered to work in human-centric areas, adapting to numerous duties and dealing with advanced actions.
Artem Sokolov, CEO and founding father of Humanoid, says: “Our mission is to create humanoid robots that carry out not solely in managed lab settings, but in addition in real-world manufacturing facility environments, dealing with significant industrial duties.
“Our collaboration with Siemens and Nvidia offers us a robust benefit by combining Nvidia’s main AI infrastructure, simulation instruments, and frameworks with Siemens’ deep industrial experience and integration capabilities. Collectively, we’ve confirmed that humanoid robots are prepared for real-world industrial deployment.”
