CS449/599 — Learning for Robotics
Each semester, students put together a final project using the tools introduced in this course. The examples shown below represent a small, curated sample that captures the diversity of approaches explored.
Fall 2024
-
Constructing tall and stable structures through sequential robotic manipulation using optimization-based planning that combines nonlinear programming with heuristic configuration search.
-
Developing an optimization-based robotic system that accurately throws objects into dynamically placed bins by combining trajectory optimization with RRT-based path planning.
-
Retrieving a target object from densely cluttered environments using a perception-driven manipulation pipeline. A multi-camera RGB-D system constructs a unified 3D point cloud, which is segmented and analyzed to prioritize object removal based on stability and hierarchy, enabling efficient grasping and placement through motion optimization.