Figure AI's humanoid robot "Jim" (F.03 model) demonstrated sustained autonomous operation in a logistics warehouse, sorting packages nonstop for over 81 hours (with livestream elements extending discussions into recent days). The robot uses onboard AI (Helix-02 neural network) to scan barcodes, recognize boxes, identify label orientations, pick/flip items, and load conveyors at speeds approaching or matching human performance (e.g., one package every ~2.9 seconds in tests). It processed tens to over 100,000 packages in extended runs without human intervention, highlighting endurance advantages over fatigued workers (no breaks needed after initial hours).
Discussions emphasize its potential for 24/7 warehouse operations amid labor shortages, though experts note challenges in precision for edge cases, exception handling, and full commercial scalability versus controlled demos. This aligns with broader shifts toward flexible AMR/humanoid integration for picking/sorting in dynamic environments. Ongoing online engagement and viewer-named milestones underscore public interest in humanoid reliability for intralogistics.