CS59200-RM: Robot Manipulation
Course Description
There have been tremendous recent advances in the capabilities of robots to achieve complex tasks, and the field sits at a precipice of broader impact; robots are moving beyond a priori engineered factory settings to un/semi-structured domains, such as moving packages around a warehouse, providing automated delivery in restaurants and hospitals, advancing human telepresence in distant locales such as underwater and in space, and more. Despite this progress (and the abundance of cool demo videos), it’s still incredibly difficult to make a robot do any one thing, let alone have a generalizable system that can handle a significant breadth of tasks in various environments. So, what’s so hard about making a robot interact with the world?
This course will answer that question and provide an in-depth understanding of the state-of-the-art in robot manipulation by surveying important landmark papers in the field as well as current recently published works. In particular, this course will have an algorithmic and computational focus, providing an understanding of the fundamental techniques necessary for manipulation. We will also cover modern advances in how statistical machine learning (particularly approaches known as deep learning, generative AI, or foundation models) are applied and used by model- and optimization-based methods to the handle the uncertainties of the real world.
Course Information
| Course Number | CS59200-RM |
| Meeting Time | TTH 1:30-2:45pm |
| Course Room | LWSN 1106 |
Resources
There is no required textbook for this course. Lecture notes, research papers, and online resources will be provided throughout the course. For the interested reader, there are recommended textbooks that cover much of the basic background necessary:
- Planning Algorithms by Steven M. LaValle
- Modern Robotics by Kevin M. Lynch and Frank C. Park
- Robotic Systems by Kris Hauser
- Robotic Manipulation: Perception, Planning, and Control by Russ Tedrake
- Automated Planning and Acting by Malik Ghallab, Dana Nau, and Paolo Traverso
A nice high-level overview of controlling manipulators is available in this blog post.
Reading Schedule
TBD