Long-horizon autonomous reasoning and operation closed-loop

Long-horizon autonomous reasoning and operation closed-loop

  • Supports CoAT task chains exceeding 30 minutes
  • Fast/slow brain collaboration: Integrated perception-reasoning-execution
  • Reinforcement learning-driven self-reflection and strategy optimization
Humanoid dexterous manipulation capabilities

Humanoid dexterous manipulation capabilities

  • Millimeter-level precision in actions like flipping cards, grasping cards, and threading rings
  • Full-hand tactile + force control system, supporting blind grasping and precise force output
  • 16-DOF exoskeleton teleoperation, highly replicating human hand movements
Generalization and multi-scenario deployment capability

Generalization and multi-scenario deployment capability

  • Achieves three-way interaction between human-robot/machine-machine/machine-environment
  • Widely applicable to complex environments such as logistics, retail, and home
  • Modular whole-machine solution, supporting large-scale low-cost deployment

Scroll to Top