About this video
A full run of the WALK model: a twelve-joint (12 DOF) quadruped built in RoboCo on Steam.
A Python script controls all four legs to walk forward and back, turn in place, side-step, and run automatic gait patterns.
The video works through the keys assigned to the WALK model in order, so you can see how the robot's motion changes with each input.
Keys in this exercise
| Key | Action |
|---|---|
| W / S | Walk forward / backward |
| A / D | Turn left / right in place |
| Q / E | Side-step left / right |
| X | Stop pose |
| Z | Relaxed pose |
| 1 | Auto mode: basic moves |
| 2 | Auto mode: walking at different heights |
| 3 | Auto mode: complex pattern |
| 0 | Relax, then end the script |
Direction keys keep the robot walking while held and return it to the default pose when released. Number keys 1, 2, and 3 run prepared automatic gait patterns.
What to watch
- The moment two diagonal legs lift and land together
- How the body's heading differs between A/D turns in place and Q/E side steps
- The short settle after you let go, until all four feet are down
Exercise details
| Book | Learning Physical AI with RoboCo: Robot Structure and Gait Control (Korean edition) |
|---|---|
| Series | Physical AI with RoboCo · QUAD-01 (WALK) |
| Chapter | PART 1 · EXPERIENCE · CHAPTER 01 |
| Related chapters | CH01·03·04·16·19~25 |
| Practice model | Book-QUAD-01-WALK |
| Python start file | q01_walk_start.py |
| Practice files version | v3.5.0 — download |
| Video ID | Q01-01-00 |
| YouTube | youtu.be/4-5uGHX3k2E (opens in a new tab) |