Free courses worth your time.
A hand-picked reading list for the path from PID loops to learned policies. Everything here is free and legally available online, and every entry is something I’d actually recommend to a friend asking where to start.
PID, root locus, Bode, Nyquist
Start here if the word 'phase margin' doesn't mean anything yet. These teach feedback from first principles — the same material every controls engineer learned in undergrad, taught well.
- Playlist
Control Systems Lectures
Brian DouglasThe friendliest on-ramp to classical control. Short, clear, animated. Start with 'What is a Transfer Function?' and just keep going.
- Playlist
Control Bootcamp
Steve BruntonState-space control, LQR, observability, controllability. The natural next step after Brian Douglas.
- Book
Feedback Systems: An Introduction for Scientists and Engineers
Åström & MurrayFree legal PDF. The standard graduate-level textbook; strong on intuition and worked examples.
LQR, MPC, trajectory optimization
Where control meets optimization. These assume you're comfortable with the classical stuff and ready to formulate control problems as constrained optimization.
- Course
Underactuated Robotics
Russ Tedrake (MIT 6.832)Open textbook + lectures + Python notebooks. LQR, trajectory optimization, policy search, running on Drake. The single best free resource on model-based robot control.
- Book
Predictive Control for Linear and Hybrid Systems
Borrelli, Bemporad & MorariSlides + notes from the Berkeley MPC course. Rigorous and comprehensive — the canonical MPC text.
- Course
Convex Optimization
Stephen Boyd (Stanford EE364A)Lectures, homework, and the free Boyd & Vandenberghe book. If you want to actually solve MPC problems, you need this.
From MDPs to policy gradients
Two tracks: the fundamentals (Sutton & Barto, David Silver) and the modern deep-RL tooling (OpenAI Spinning Up, CS285). Do at least one from each.
- Book
Reinforcement Learning: An Introduction (2nd ed.)
Sutton & BartoFree legal PDF. The foundational text. Read the first ten chapters before any deep-RL course.
- Lectures
RL Course
David Silver (DeepMind)The classical companion to Sutton & Barto — ten lectures, taught by one of the authors of AlphaGo. Slides and videos free.
- Course
Deep Reinforcement Learning (CS285)
Sergey Levine (Berkeley)The modern deep-RL curriculum. Policy gradients, actor-critic, model-based RL, offline RL, imitation learning. Lectures and assignments free.
- Notes
Spinning Up in Deep RL
OpenAIA pragmatic, code-first bridge between the theory courses and actually training agents. Clean PyTorch implementations of the core algorithms.
- Course
Reinforcement Learning (CS234)
Emma Brunskill (Stanford)Strong alternative to CS285 with more emphasis on theory and exploration. Lecture videos available.
Manipulation, locomotion, learning for control
Where the rubber meets the road. These courses assume control and ML foundations and focus on actually getting things to move.
- Course
Robotic Manipulation
Russ Tedrake (MIT 6.4210)Companion to Underactuated. Grasping, kinematics, pose estimation, motion planning, and — increasingly — learned policies. Free book + labs.
- Lectures
Introduction to Robotics
Oussama Khatib (Stanford CS223A)The classical robotics sequence: transforms, kinematics, Jacobians, dynamics. Dry but complete.
- Course
Deep Learning for Robotics (CS237B / CS237)
StanfordModern robotics from an ML angle: imitation learning, VLAs, diffusion policies. Newer, less stable, worth skimming.
- Lectures
Legged Robots: Reinforcement Learning Workshops
ETH Zurich / VariousCurated lecture materials on sim-to-real RL for legged locomotion — the actual recipes used in recent parkour / humanoid papers.
Linear algebra, probability, optimization
If any of the above feels like hieroglyphs, back up. These three courses cover the math that control and ML both sit on.
- Course
Linear Algebra (MIT 18.06)
Gilbert StrangThe canonical free linear-algebra course. Do the problem sets.
- Course
Introduction to Probability
Blitzstein (Harvard Stat 110)Intuitive probability from scratch. Free lectures + textbook.
- Book
Mathematics for Machine Learning
Deisenroth, Faisal & OngFree PDF. Linear algebra, calculus, probability, optimization — aimed exactly at the ML audience.