The full curriculum
Six months, one page
119 lectures, about 95 hours of video, 6 projects. Roughly 2 hours a day: one of lectures, one of code.
M0
optional
Python and your setup
Only if you've never coded. Skip it if you already write Python.
The Python Tutorial + MuJoCo tutorial notebook
~4 weeks, 1 h a day
done0/6 · 0%M1
The math that everything else stands on
Vectors, matrices and least squares, taught as tools, not proofs.
ENGR108
~18.5 h of video + practice
done0/7 · 0%M2
The robot's body
How an arm moves, where its hand is, and how much force its motors need.
CS223A
~19 h of video + practice
done0/8 · 0%M3
How a robot sees and senses
Cameras, IMUs and lidar, and the neural networks that read them.
CS231N + EE259
~17 h of video + practice
done0/7 · 0%M4
Planning and control
How a robot decides what to do next, when it knows how the world works.
AA203
~15.5 h of video + practice
done0/8 · 0%M5
Teaching robots instead of programming them
Imitation learning and reinforcement learning, from Chelsea Finn and the AA203 team.
AA203, lectures 13 to 19 + CS224R
~17 h of video + practice
done0/9 · 0%M6
The frontier, and your first robot
Talks from the people building robot AI, and one project that proves everything.
9 talks & seminars
~8.5 h of talks + final project
done0/9 · 0%
How to actually finish this
- 01Watch at 1.25x. Most lectures stay easy to follow, and it saves about 19 hours over the whole path.
- 02Never watch more than two lectures without writing code. The practice tasks are the course. The videos are support.
- 03Keep the free textbooks open next to the video. When a professor goes too fast, the book usually explains it slower.
- 04Put every project on GitHub from day one, with a README about what broke and how you fixed it.
- 05Budget 2 hours a day: 1 hour of lectures, 1 hour of code. Keep your job while you do this.