X Square Robot Unveils Embodied AI Workflow at WRC 2026

The company’s exhibit centers on the WALL-B model, which powers six-axis robotic arms now capable of processing 1,816 parcels per hour with 98% accuracy. These systems autonomously handle variable package sizes and materials, including delicate label flattening, mimicking human-like adaptability in industrial settings. Beyond the warehouse, X Square is pushing into domestic spaces through its X Family Member Program, which integrates robots into households for daily tasks like cleaning and meal assistance.

Demonstrations at the event also highlight the company’s progress in fine motor control and semantic reasoning. In one test, robots responded to natural-language requests to arrange flowers, adjusting their physical movements in real-time as object positions shifted. These capabilities are underpinned by the QUANXTA Zero platform, which allows for the production of training data without the need for a physical robot body. By linking this data pipeline directly to dexterous-hand manipulation and foundation models, X Square aims to bridge the gap between abstract machine learning and the unpredictable physical world.