Learn · MIT OpenCourseWare

A photograph of a flying robot placed above a photograph of a legged robot.A photograph of a flying robot placed above a photograph of a legged robot.

Course6.83223 lectures28 h 28 min

Underactuated Robotics

Prof. Russell Tedrake · Massachusetts Institute of Technology

Start with lecture 1 At MIT OpenCourseWare

About

Robots today move far too conservatively, using control systems that attempt to maintain full control authority at all times. Humans and animals move much more aggressively by routinely executing motions which involve a loss of instantaneous control authority. Controlling nonlinear systems without complete control authority requires methods that can reason about and exploit the natural dynamics of our machines.

This course discusses nonlinear dynamics and control of underactuated mechanical systems, with an emphasis on machine learning methods. Topics include nonlinear dynamics of passive robots (walkers, swimmers, flyers), motion planning, partial feedback linearization, energy-shaping control, analytical optimal control, reinforcement learning/approximate optimal control, and the influence of mechanical design on control. Discussions include examples from biology and applications to legged locomotion, compliant manipulation, underwater robots, and flying machines.

Acknowledgments

Professor Tedrake would like to thank John Roberts for his help with the course and videotaping the lectures.

Lectures

Video Lectures

  1. Lecture 1: Introduction 1 h 15 min · Video
  2. Lecture 2: The Simple Pendulum 1 h 9 min · Video
  3. Lecture 3: Optimal Control of the Double Integrator 1 h 17 min · Video
  4. Lecture 4: Optimal Control of the Double Integrator (cont.) 1 h 25 min · Video
  5. Lecture 5: Numerical Optimal Control (Dynamic Programming) 1 h 13 min · Video
  6. Lecture 6: Acrobot and Cart-pole 1 h 21 min · Video
  7. Lecture 7: Swing-up Control of Acrobot and Cart-pole Systems 1 h 6 min · Video
  8. Lecture 8: Dynamic Programming (DP) and Policy Search 1 h 14 min · Video
  9. Lecture 9: Trajectory Optimization 1 h 10 min · Video
  10. Lecture 10: Trajectory Stabilization and Iterative Linear Quadratic Regulator 1 h 20 min · Video
  11. Lecture 11: Walking 1 h 16 min · Video
  12. Lecture 12: Walking (cont.) 1 h 12 min · Video
  13. Lecture 13: Running 58 min · Video
  14. Lecture 14: Feasible Motion Planning 1 h 14 min · Video
  15. Lecture 15: Global Policies from Local Policies 1 h 19 min · Video
  16. Lecture 16: Introducing Stochastic Optimal Control 1 h 24 min · Video
  17. Lecture 17: Stochastic Gradient Descent 1 h 17 min · Video
  18. Lecture 18: Stochastic Gradient Descent 2 1 h 19 min · Video
  19. Lecture 19: Temporal Difference Learning 1 h 20 min · Video
  20. Lecture 20: Temporal Difference Learning with Function Approximation 1 h 18 min · Video
  21. Lecture 21: Policy Improvement 1 h 16 min · Video
  22. Lecture 22: Actor-critic Methods 1 h 11 min · Video
  23. Lecture 23: Case Studies in Computational Underactuated Control 1 h 3 min · Video

Syllabus and materials

At MIT OpenCourseWare, as the course publishes them.

Credits and licence

Institution
Massachusetts Institute of Technology
Department
Electrical Engineering and Computer Science
Teacher
Prof. Russell Tedrake
Term
Spring 2009
Level
Graduate
Language
English
Subjects
Computer science, Engineering
Source
MIT OpenCourseWare
Licence
CC BY-NC-SA 4.0, as MIT OpenCourseWare states it. Non-commercial use only. MIT OpenCourseWare's terms
Attribution
Prof. Russell Tedrake. Underactuated Robotics. Spring 2009. Massachusetts Institute of Technology: MIT OpenCourseWare, https://ocw.mit.edu/. License: Creative Commons BY-NC-SA.
Who can watch
MIT OpenCourseWare gives no rating. Invisible Network shows it to Tweens profiles and older, judged from what the source says about it. How courses are placed