Week 02
Week 2 Summary: Kinematics & Coordinate Systems Complete
Published: 2026-04-12 | Author: Smartotics Learning Journey | Reading Time: 6 min
Figure 1: Week 02 Complete progress dashboard showing 7 days and curriculum overview
Table of Contents
Week 2 Complete!
We’ve covered coordinate systems, rotation matrices, homogeneous transforms, Euler angles, quaternions, and forward kinematics — the mathematical foundation that everything else in robotics builds upon. This week marks the transition from “what is a robot” to “how do we compute with robots.”
Week 2 Overview
| Day | Topic | Duration | Status |
|---|---|---|---|
| Day 08 | Coordinate Systems & Frames | 30 min | Completed |
| Day 09 | Rotation Matrices | 30 min | Completed |
| Day 10 | Homogeneous Transform Matrices | 35 min | Completed |
| Day 11 | Euler Angles vs Quaternions | 30 min | Completed |
| Day 12 | Forward Kinematics (Theory) | 35 min | Completed |
| Day 13 | Python Practice: Kinematics | 45 min | Completed |
| Day 14 | Week 2 Summary & Review | 20 min | Completed |
Day-by-Day Key Learnings
Day 08: Coordinate Systems
Every robot uses multiple frames: World (fixed), Body (robot’s own), Tool (end-effector), Joint (each motor). Transforms between frames are the foundation of everything.
Day 09: Rotation Matrices
3x3 matrices that encode orientation. R-1 = RT (the transpose). Three basic rotations (X, Y, Z) combine into any 3D rotation via matrix multiplication.
Day 10: Homogeneous Transforms
4x4 matrices combining rotation + translation. Chain them: T_0_n = T_0_1 * T_1_2 * … * T_(n-1)_n. This is forward kinematics.
Day 11: Euler Angles vs Quaternions
Euler angles (roll/pitch/yaw) are intuitive but suffer from gimbal lock. Quaternions (4 numbers) are the industry standard — no gimbal lock, smooth interpolation (SLERP), computationally efficient.
Day 12: Forward Kinematics
Given joint angles, calculate end-effector pose using D-H parameters and chained transform matrices. The math works for any number of joints.
Day 13: Python Practice
Built a working kinematics library from scratch: rotation matrices, homogeneous transforms, D-H forward kinematics, workspace visualization. Run the code and see your robot move.
Core Concepts to Remember
The Robot Kinematics Toolbox
| Tool | Format | Use When |
|---|---|---|
| Rotation Matrix | 3x3 matrix | Mathematical derivations, transformations |
| Homogeneous Transform | 4x4 matrix | Chaining multiple transformations (FK) |
| Euler Angles | 3 numbers (φ,θ,ψ) | Human input/output, IMU readings |
| Quaternions | 4 numbers (w,x,y,z) | Internal calculations, interpolation, ROS |
Key Formulas
Rotation about Z (memorize the pattern)
R_z(θ) = [[cosθ, -sinθ, 0], [sinθ, cosθ, 0], [0, 0, 1]]
Homogeneous Transform
T = [[R_3x3 | d_3x1], [0 0 0 | 1 ]]
Inverse Transform (fast!)
T_inv = [[R.T | -R.T * d], [0 0 0 | 1 ]]
Chain: base to end-effector
T_total = T_0_1 @ T_1_2 @ T_2_3 @ … @ T_(n-1)_n
What’s Next: Week 3 Preview
Week 03: Inverse Kinematics
FK is the easy direction. Inverse Kinematics (IK) is the hard problem: “Given where I want the end-effector, what joint angles do I need?” This is the computation that runs 100+ times per second every time you drag a robot in a simulation.
| Day | Topic |
|---|---|
| Day 15 | Inverse Kinematics: The Hard Problem |
| Day 16 | Geometric Solutions (2-DOF, 3-DOF) |
| Day 17 | Algebraic Solutions |
| Day 18 | Numerical Solutions (Jacobian Method) |
| Day 19 | Pieper’s Criterion & Analytical Solutions |
| Day 20 | Multiple Solutions & Optimization |
| Day 21 | Week 3 Summary |
Prerequisite check: Make sure the Python code from Day 13 (forward kinematics) works before starting Week 3. You’ll build on it.
Personal Reflection
What Worked This Week
- Progressive depth: Frames → Rotation → Transforms → FK builds naturally
- Hands-on code: Day 13 Python practice made abstract math concrete
- Real-world examples: FANUC pick-and-place, gimbal lock in Apollo 11
Challenges
- Math density: This week was the most math-heavy so far
- Matrix multiplication: If this is rusty, review 3Blue1Brown’s linear algebra series
Moving Forward
Week 2 is the most mathematics-heavy single week in the entire curriculum. After this, the math stays practical and always connects to code or hardware. If you can compute forward kinematics, you can handle everything that follows.
“The mathematics of kinematics is beautiful because it turns abstract numbers into physical movement. When your code makes a robot arm reach to the right spot for the first time, the math stops being abstract.”
FAQ
Q: Should I go back and review any of Week 1?
Not unless you struggled. Week 2 builds on Week 1’s robot classification and components. If you remember Sense-Think-Act and the 3 robot categories, you’re fine.
Q: The Python code looks complex. Do I need to memorize it?
No. Memorize the concepts (what a rotation matrix does, why we chain transforms). The code is available in PythonRobotics (github.com/AtsushiSakai/PythonRobotics). But do write it once from scratch so you understand it.
Q: I’m stuck on rotation matrices. Can I skip ahead?
Don’t. Week 3 (Inverse Kinematics) requires Week 2. If rotation matrices are unclear, watch 3Blue1Brown’s linear algebra and Brian Douglas’s videos again — they’re the best visual explanations available.
Key Takeaways
Disclaimer
For educational purposes only. This article is part of a structured learning curriculum.
Image Credits: AI-generated illustrations. © 2026 Smartotics Learning Journey.