PyBullet Tutorial (TA Session)

This is a PyBullet tutorial for CS4278/CS5478 Intelligent Robots: Algorithms and Systems. This repository contains a step-by-step introduction to PyBullet for robotics simulation. The tutorial is structured into six example scripts. It covers basic simulation, loading robots, forward & inverse kinematics, control, and a full pick-and-place demo with the Franka Panda arm.

---

📂 Files Overview

0. Hello PyBullet — 000_hello_pybullet.py

• Minimal example: connect to PyBullet, load a plane and a falling cube, step the simulation.

• Teaches how to:

  • Start/stop simulation
  • Load URDF assets
  • Get object states (getBasePositionAndOrientation)

---

1. Load a Robot — 001_get_a_robot.py

• Load the Franka Panda arm and keep it fixed at the base.

• Demonstrates:

  • Inspecting joints
  • Disabling default motors
  • Understanding fixed base vs free base

---

2. Forward Kinematics (FK) — 002_forward_kinematics.py

• Visualize robot link frames using a small URDF marker.

• Demonstrates:

  • getLinkState to get a link’s world transform
  • Attaching markers to links for FK visualization
  • Understanding Panda’s link hierarchy

---

3. Inverse Kinematics (IK) — 003_inverse_kinematics.py

• Add GUI sliders for an interactive IK demo.

• Demonstrates:

  • calculateInverseKinematics
  • Using GUI sliders to set EE target pose
  • Comparing target vs current EE frame

---

4. PD Control — 004_pd_control.py

• Replace direct joint resets with PD controllers.

• Demonstrates:

  • POSITION_CONTROL mode
  • Effects of P/D gains and torque limits

---

5. Pick and Place — 005_pick_and_place.py

• Full pipeline: detect cube, compute grasp poses, move Panda with IK+PD, control gripper, place cube at a target.

• Features:

  • Random red cube dropped in front of robot
  • Random green cube as goal (visual only)
  • Keyposes (pregrasp, grasp, lift, place) visualized with markers + labels
  • Gripper open/close with prismatic joints
  • Helpers for joint limits, object settling, and smooth IK

---

🛠️ Setup

1. Install dependencies:

   pip install pybullet numpy

2. Clone this repo and run examples:

   python 000_hello_pybullet.py
   python 001_get_a_robot.py
   ...
   python 005_pick_and_place.py

3. Assets:

   • Place additional URDFs under assets/:

     • assets/plane/plane.urdf
     • assets/franka_panda/panda.urdf
     • assets/frame_marker/
     • assets/cubes/5cm_green_cube.urdf, 5cm_red_cube.urdf

---

References

• PyBullet Quickstart Guide.
• PyBullet Examples.