I am currently a 2nd year Masters student in the Mechanical Engineering Department of the University of Illinois at Urbana Champaign. I am a graduate research assistant in the Reliable Autonomy Group @ UIUC. I enjoy full stack robotics projects, working with both software and hardware.
In my free time I love to build and ride bikes, and apply endless upgrades to my ender3 printer.
Thingiverse CAD Designs: https://www.thingiverse.com/byang11259/designs
Past Projects
Vision based drone Racing
Fall 2024
Work in progress… GitHub Link
ME461 Project: Segbot
Fall 2024
C program implementation of balancing segbot on Texas Instruments LAUNCHXL-F28379D.
UAV Simulation: Vision-based Closed Loop Navigation
Fall 2023
Closed loop vision-based simulation for quadcopter. Images generated from NeRF model of the flying arena. Quadcopter running monte carlo particle filter localization.
TACAS Submission 2024
GitHub Link
UAV Landing Using Perception Contracts
Spring 2023
Demonstrate learning-based approach that can automatically characterize the error of a perception module from data and use this for safe control. The proposed approach constructs a perception contract (PC), which generates a set that contains the groundtruth value that is being estimated by the perception module, with high probability.
The PC technique is applied for the vision pipeline in the UAV safe-landing scenario. A PC is trained from ground truth data. A controller is designed to utilize the PC. Experiments show that with the control algorihtm with learned PC can safely land the quadcopter despite error from the perception module, while the baseline algorithm without PC failed to do so.
Accepted ICRA 2024
ME446 Project: 6DOF Manipulator Scripting
Spring 2023
Complete the test environment while demonstrating task space PD and feed forward control tracking, force control, and impedance control.
Automated Needle Insertion Mechanism for Needle Tribology
Fall 2022
Needles are subject to a variety of forces as they are inserted into a body. Through a better understanding of the forces present as needles are traveling through soft material, people can optimize the longevity of needles and reduce its damage on patients.
The purpose of this project is to construct a setup that provides a controlled and repeatable study of the needle insertion process. A mechanism mounts on top of an inverted microscope to image the contact in situ and collects compression and tension forces. As a senior design project, the project has been commissioned and directed by Dr. Alison Dunn for future research within her lab.
ECE484 Project: Carla Simulator
Fall 2022
The simulation is part of the Generalized Racing Intelligence Competition (GRAIC). Given ground-truth perception information, the focus of the project is on the planning and control of the vehicle. An aggressive and fast controller is developed to complete the generalized race course in the shortest amount of time and evade the most number of obstacles.
The controller utilizes PRM-variant as its sampling strategy and finds shortest path by performing topological sort in a directed acyclic graph.
Lateral control of the car is based on a PD controller. For longditudinal control, the acceleration is based on path curvature. Using the current position of the car and the next two waypoints on our path, we calculated the radius of the circle generated from those three points. The radius of the circle tells us how much curvature is coming up in our path. We tuned a logarithm relationship between the radius and desired vehicle speed
Design and Clinical Validation of a Robotic Ankle-Foot Simulator for Ankle Clonus
Spring 2020
Undergraduate Researcher at Human Dynamics and Controls Laboratory (HDCL), advised by Dr. Pei Yinan and Professor Elizabeth Hsiao-Wecksler
The standard clinical assessment of neurological diseases lies with a clinicians’ manual techniques to elicit muscle response. Their effectiveness and accuracy largely lies in their past experience and training. Traditional clinical training of motor skill assessment is limited because it relies on the instructor bringing in live patients for students to practice upon. The project aims to provide a medical training simulator that will be accessible and provide consistent training opportunities for students that replace or reduce the need for human patients .
Design a support system for the leg joint that will provide:
- Ability to imitate situations when patient is lying down and their legs flat out, and when patient is sitting down
- Ease of Transportation: The total weight of the frame is approximately 1.4 kg. The base HDPE plate is removable with 4 winged nuts for a more compact carry
- Ability for practitioner’s secondary hand to apply leverage to the underside of the thigh segment
Passive Hydraulic Simulator for Biceps Spasticity
Fall 2019
Undergraduate Researcher at Human Dynamics and Controls Laboratory (HDCL), advised by Dr. Pei Yinan and Professor Elizabeth Hsiao-Wecksler
Spasticity is often found in patients with stroke, spinal cord injuries, and other neurological disorders causing abnormal motor activity. Spasticity can cause involuntary muscle contractions, spasms, and continuous muscle stiffness that affect body growth.
The current assessment for spasticity relies on the past experiences of the clinician, which is difficult in training due to the difficulty in bringing in live patients. The motivation of this project is to provide a self-contained and unpowered simulator that will provide additional hands-on practice in the current training program.
The simulator will replicate the response of spasticity patients through an adjustable hydraulic piston. When certain technique checkpoints are accomplished, the electronic feedback will inform the user on their results.