A recent X post and associated coverage highlight the demonstration of the RobotEra L7 humanoid robot performing autonomous package sorting. Powered by the ERA-42 Vision-Language-Action model, the L7 uses onboard vision for real-time environmental understanding and decision-making without manual reprogramming. Demonstrated in logistics environments and referenced in connection with WAIC 2026 activity, the system identifies, scans and handles parcels with precision, illustrating progress toward flexible, humanoid-form automation capable of operating in unstructured or semi-structured warehouse and sorting settings.
Complementary earlier reports describe broader commercial deployments of the same platform across multiple Chinese logistics centers operated by major carriers, with units achieving high relative efficiency on feeding and sorting tasks while running continuously. The combination of multimodal sensing, dexterous manipulation and fleet-level coordination positions such systems as potential bridges between traditional fixed automation (conveyors, AS/RS, conventional AMRs) and the remaining human-dependent flexible handling steps.