This course covers the fundamentals of robot manipulators:

• Forward and inverse kinematics using the product of exponentials (PoE) formula
• How to design a robot
• How to run robot simulations

Final project: RB5 pipetting. In teams of two, students program a Rainbow Robotics RB5 cobot in a MuJoCo simulation to do a pipetting task. The project has three parts, each building on the one before:

• Forward kinematics: derived by hand in NumPy and checked against MuJoCo
• Inverse kinematics: the Jacobian and an IK solver that move the robot through the task
• Pipetting hand: a hand design that grasps the pipette, presses its bulb, and returns it reliably despite errors in the pipette holder's position, evaluated in simulation and designed in 3D CAD

The reference video below shows what each part looks like when it works.

This course covers how to build and control real robot hardware:

• Designing PCBs (KiCad lecture videos on YouTube)
• Motor basics (CAN, CANopen, PWM, Transmission)

In parallel, we explore the enabling technologies and fields of robotics by reviewing the Springer Handbook of Robotics. Our chapter reviews are collected on the Notion page below.

This course covers how robots learn to act:

• Imitation learning
• Reinforcement learning, including PPO and TRPO
• Robot foundation models, such as Vision-Language-Action (VLA) models