The goal of this project was to track and predict the trajectory of a ball as it bounces around a room.  The 3-D location of the ball was observed using an RGB-D (Red Green Blue & Depth) camera, specifically the Microsoft Kinect V2 sensor.  Tracking the ball required a few steps.  First we generated an algorithm to detect the ball in each depth frame.  Then we used and Extended Kalman Filter to predict the location of the ball when it failed to be identified in a frame as well as predict the future location of the ball. 

Collaborators: J. Calderon, J. Lee

For more information please visit: PACT

 In this project we used a LQR (Linear Quadratic Regulator) in order to have a simple model of a Bipedal robot follow a stable trajectory.  Additionally, we used a feedforward controller to improve tracking. 

Collaborators: K. Ueunten

Bipedal Project Writeup

The model and desired trajectory were adapted from: Feedback Control of Dynamic Bipedal Locomotion

 I designed and produced a latching mechanism that allows a quad-rotor to pick up and release a ground vehicle.  This was part of a larger project to have a quad-rotor autonomously transport ground vehicles.

Collaborators: E. Schoof, B. Oakes, G. Johnson

I worked on this project in my spare time in order to gain experience working with the Linux (Ubuntu) and ROS operating systems.  I chose to implement SLAM (simultaneous localization and mapping) in simulation and on a TurtleBot hardware testbed.  Although this work mostly consisted of implementing existing software, it allowed me to gain familiarity with the very popular and useful ROS OS.

For more information please visit:  Zero to Slam