Gallery · RoboDojo

stack blocks, layout 9

Task: “Stack the three blocks with different textures.” DeepSeek-V4.1-Flash, code mode: 7 decisions, 27 motions, solved.

2× · 21 s of motion
  1. 1 of 7 · step1.py
  2. 2 of 7 · step2.py
  3. 3 of 7 · step3.py
  4. 4 of 7 · step4.py
  5. 5 of 7 · step4b.py
  6. 6 of 7 · step5.py
  7. 7 of 7 · step6.py
  1. grasp_execute
  2. grasp_execute
  3. grasp_execute
  4. grasp_execute
  5. grasp_execute
  6. move_pose
  7. set_gripper
  8. move_pose
  9. move_pose
  10. grasp_execute
  11. grasp_execute
  12. grasp_execute
  13. grasp_execute
  14. grasp_execute
  15. move_pose
  16. set_gripper
  17. move_pose
  18. move_pose
  19. grasp_execute
  20. grasp_execute
  21. grasp_execute
  22. grasp_execute
  23. grasp_execute
  24. move_pose
  25. set_gripper
  26. move_pose
  27. return_home
Decisions

The recording is the robot's own cameras at 2×; the time between motions, while the model thinks, is not filmed. The 3D replay is the camera's picture in depth after each decision, with the paths it planned, to turn and step through.