100 Days of Robotics

I am making this curriculum while I go. I keep in mind about people like me who did not come from robotics or engineering background, but with basic maths and enough of python basics, so that one can learn the fundamentals and get a grasp by hands-on practice.

There will be days which would look like can be done in one day but actually because of life or coding errors would take more than 1 day but that is fine! I am hoping to make this challange as something to teach and learn for the sake of love for it and not to impose artificial deadlines!

I will be laying this in 6 phases, namely,

After I have update at least the Phase 1 fully, then I will release the github repo for everyone to use as reference. Meanwhile you can check on my timely updates in my twitter

Phase 1: Foundation & Simulation environment

D1 - first robot
>> install all python and dependencies.
>> install PyBullet
>> define a simple robot with URDF
>> give basic movement functionality
>> simulate it.

D2 - programming for robots
>> Know basic functional python
>> know how URDF works, joints and links, etc.
>> after going through documentations to check how to move the joints like setJointMotorControl2 code the movements
>> make the robot dance smoothly in a pattern

D3 - robot simulation environment
>> install MuJoCo
>> tinker with mujoco models formatted in xml
>> learnt how to work with mujoco xml; coordinates, movements, etc.
>> setup complex obstacle environment in mujoco
>> custom WASD movement with mujoco

D4 - physics in simulation
>> understand what physics engines are
>> tweak around the basics of gravity and collision with BOX2D
>> in MuJoCo, experimented with different gravity
>> experimented collision between a sphere and a squishy cube

D5.1 - learn the theory of representation
>> Learnt about system presentation
>> learnt about system translation and rotation in simple states
>> learnt to apply rotation in 3d space
>> learnt about homogenous transformation to trivialize transformation in higher spaces.
>> studied the euler angles and euler rotation theorem

D5.2 - theory for forward and inverse kinematics
>> used tkinter to visualise a 2D forward kinematics principle
>> learnt all the necessary inverse kinematics and joint angles
>> learnt how to solve for IK with both trigonometric and intersection of circle methods
>> created a dashboard with tkinter to update the arm with given end point co-ordinate

D6 - Intro to Robot Sensors and obstacle detection
>> learn pygame and pymunk to simulate
>> study collision detection theory - Separated Axis Theorem
>> simulate collision detection
>> add bounce back path after each collision

D7 - Open loop basic robot control
>> setup webots
>> learn about open loop
>> learn simple drive system mechanics for e-puck bot
>> implemented a time based predetermined path to follow

D8 - Closed loop feedback control
>> closed loop control and feedback
>> introduction to PID controllers
>> understand and fine tune kp, ki, kd
>> implement a feedforward control for a robot to move to a box but never collide

D9 - Tuning PID controllers
>> learnt about PID tuning- heuristic, Ziegler-Nichols, Cohen-Coon.
>> learnt how the nodes in webots work, while implementing a track path
>> tested and analysed a lot of kp values for ZN methods to calculate TU and KU for line follower
>> tested a lot on webot, external controls and analysing time v/s control curves
>> finally heuristically achieved control for line following with only two sensor
>> implemented Ziegler-Nichols tuning method perfectly.

D10 - Forward Kinematics on UR3E
>> Review forward kinematics
>> Revised homogeneous transformation for rotations and translations
>> learnt about Denavit-Hartenberg notation
>> implemented manual control for UR3E bot
>> learnt about importance of choosing axis when performing homogeneous transforms
>> real time FK calculator with tkinter