For this project, I created a motion planning and control interface for a portable, two-armed desktop robot for animal test subject experiments. This project was conducted in collaboration with the Sensory and Neural Systems Engineering Group. I combined two WidowX 200 arms into a shared motion planning and control interface that was fully integrated with ROS. Users could generate sequences of dual arm poses with end effector and joint constraints to present sets of stimuli to rodents without the arms colliding with one another. A zero-gravity feature enabled users to manually configure the arms to different poses and save them. I wrote the motion planning system in C++ and also created a separate Python API. This system is being used for a project in which an autonomous system must bend a flexible wire into a tool that can be used to perform some task.

For this project, our team programmed a Baxter robot to play a game of Yahtzee against a human opponent. We used a graph-based object detection algorithm to detect dice on a playing surface. We coupled our object detection algorithm to a motion planning scheme, which was used to grasp the dice and transport them to a nearby cup. The Baxter robot could then pick up the cup and roll the dice onto the playing surface.

Team members: Feiyu Chen, Vikas Gupta, Zhicheng Zhang, Milli Schlafly.

For my senior design project at UIUC, my team and I created mobile robot cart for transporting bins around warehouses. This project was conducted with John Deere as part of their effort to modernize their warehouses. We implemented an april tag detection system to estimate the distance from the robot cart to a designated work shelf drop off point. We used a linear servo and combined stepper motor/lead screw system to push bins containing parts onto the designated shelf at the correct distance. We also created an electronic gating system to keep the bins on the shelf while the robot cart was in motion.

Team members: Zixu Zhang, Krishna Jayaraj, Jiaxi Wen, Taehyun Park.