Aditya Mandalika (adityavk AT cs.uw.edu)
MW 3:00 - 4:20 PM, MGH 295
At the crossroads of robotics, artificial intelligence and algorithms, this graduate level course delves into the theory and algorithms that enable robots to physically interact with and manipulate their environment. We will begin with a primer on the representation of the planning problem, and the geometry of manipulation configuration space. We will then discuss state-of-the-art search and motion planning algorithms, their theoretical guarantees, and their computational requirements and limitations. You will also learn how to address clutter and uncertainty in manipulation tasks.
In practice, planning problems necessitate the application of a mix of algorithms. By the end of this class, you will learn to describe, compare, and systematically choose the set of algorithms that best solve the planning problem presented.
Grading will be based on homework assignments and class participation. The assignments will be designed in Python.
All resources are posted on Canvas.