Saad Ullah
⬤ Chinese robotics company AgiBOT (Zhiyuan) just dropped its newest humanoid robot aimed at retail and logistics automation. The company unveiled Genie2—a wheeled humanoid built specifically for instant retail fulfillment centers—right after announcing they'd mass-produced their 5,000th humanoid robot. Photos from the launch show Genie2 already working in structured warehouse settings, handling shelving systems and interacting with automated equipment.
⬤ Genie2 handles pretty much the full workflow you'd expect in a retail fulfillment center. It identifies online orders, maps out navigation routes, picks products, completes orders, and even restocks shelves automatically. The robot can grab items made from different materials, so it's versatile enough to work across various product types. Multiple Genie2 units can operate simultaneously as order pickers and shopping assistants, which could seriously boost automation levels in retail facilities.
⬤ The robot came out of a partnership between AgiBOT and Shanghai startup ACE Robotics (Daxiao). AgiBOT brings the humanoid hardware to the table, while ACE Robotics provides real-world datasets that help Genie2 better understand its environment. This data powers Genie2's vision-language-action model, letting it handle tasks autonomously through the entire shelf-to-shelf process. The system also plugs into existing business platforms, making it compatible with current retail and inventory management setups.
⬤ What's interesting here is the shift toward more adaptable robotics beyond fixed automation. ACE Robotics is working on building a general-purpose robotic "brain" that could run on multiple robot types—humanoids, robotic dogs, you name it. AgiBOT also mentioned deploying real-world reinforcement learning systems in pilot manufacturing environments, tackling the challenge of getting robots to adapt and make decisions on their own. These developments signal a move toward flexible embodied AI that can handle complex, end-to-end tasks in changing retail and industrial spaces, potentially reshaping how fulfillment and manufacturing operations scale with intelligent robotics.
Saad Ullah
