COMP 765: Intelligent Robotics, Winter 2020
Instructor
david.meger@X
Office: McConnell 112N
Office Hours: Tuesdays after class (10-11am)
Dave's office MC112N
X = mcgill.ca
-------------------------------------
Teaching Assistant
Raihan Seraj
raihan.seraj@Y
Office: McConnell 438
Office Hours: Wednesdays (10:30-11:30am)
Y = mail.mcgill.ca
News
- Feb 27, 2020: We'll start to assign papers for after Reading Week using this list: (online>
- Jan 7, 2020: Welcome to the course!
Overview
COMP 765 is a research seminar on Intelligent Robotics. We will see how classical ideas starting from Gauss and Newton intersect with the latest machine learning on-board systems that can sense the world and act upon it. The class will begin with lectures on definitional problems and algorithms in robotics. We will then transition to mixed student-lead and instructor-lead discussions of recent developments in research and in practice. The emphasis is on algorithms, probabilistic reasoning, learning to improve behaviors using data, and decision making under uncertainty, as opposed to electromechanical systems design. We will broadly cover the following areas:
- Models for the geometry and motions of robots
- Simulation to predict movements over time
- Control that produces desired behavior
- Planning safe and effective paths through the environment
- Decision making under uncertainty which unites perception with planning and control
- Learning to perceive the world and predict motions
- Interaction including learning from human demonstrators and working with teams of humans and robots
Assignments
- Assignment 1: Dynamics and Control (pdf)
Schedule
Lectures will be on Tuesdays and Thursdays in McConnell Engineering Room 103, 8:35-9:55am each week from Jan 7th until April 9th except March 2-6, McGill's study break.Week | Topics | Slides | References |
---|---|---|---|
1 |
Introduction Kinematics and Dynamics |
Lecture 1 - (pdf) (pptx) Lecture 2 - (pdf) (pptx) |
PR Chapter 1. Craig Chapters 1-4. |
2 | Optimal Control Formulation LQR and DDP |
Lecture 3 - (pdf) (pptx) Lecture 4 - (pdf) (pptx) | |
3 | Optimal Control Formulation and algorithms |
Lecture 5 - (pdf) (pptx) Lecture 6 - (pdf) (pptx) | |
4 | Planning |
Lecture 7 & 8 - (pdf) (pptx) | |
5 | Perception and Estimation Probabilistic estimation: particle and Kalman filters |
Lecture 8 - (pdf) (pptx) Lecture 9 - (pdf) (pptx) | |
6 | SLAM EKF, graph SLAM, visual navigation |
Lecture 10 - (pdf) (pptx) | |
7 | Decisions Under Uncertainty POMDP, black-box optimization for robotics, coverage and exploration |
Lecture 11 - (pdf) (pptx) Lecture 12 - (pdf) (pptx) | |
8 | Model Learning Model-based RL, Gaussian Processes for robotics, black-box optimization |
Lecture 14 - (pdf) (pptx) | Probabilistic Inference and Learning for Control (PILCO) GP-BayesFilters |
9 | Imitation Learning and Interaction Inverse Optimal Control, active imitation, Dagger. Multi-agent systems of robots and humans. |
||
10 | Student project proposals |
||
11 | Research talks and papers Presented by students and guests |
||
12 | Research talks and papers Presented by students and guests |
||
13 | Project presentations Final demos and show-and-tell |
Marking scheme
- Assignments- 20%
- Midterm - 10%
- Research paper presentations - 20%
- Project - 50% split between proposal, final presentation and final report
Textbooks
These are all optional, but give great background on the subject. We will sometime assign readings from the material that is available freely on the web.- Planning Algorithms, by Lavalle (PLAN in outline)
- Probabilistic Robotics, by Thrun, Fox, and Burgard (PR in the outline)
- Introduction to Robotics, Dynamics and Control, by Craig (Craig in the outline)
- Gaussian Processes for Machine Learning, by Rasmussen and Williams (GPforML in the outline)
- Computational Principles of Mobile Robotics, 2nd edition, by Dudek and Jenkin
- Robotics, Vision, and Control, by Corke
- Introduction to Autonomous Mobile Robots, by Siegwart, Nourbakhsh, Scaramuzza