2023
Learning to Manipulate with Scene Graph
May 2023 – Jun 2023, Carnegie Mellon University, Pittsburgh, PA
- Proposed scene graph representation interprets instructional videos as manipulations of object attributes and relationships
- Enable RL agents to learn from a single YouTube video and perform tasks in significantly different environments, showing promising results for real-world applications
2022
Imitation of Pushing from Video Demonstrations
Nov 2022 – Dec 2022, Carnegie Mellon University, Pittsburgh, PA
- Implemented Kalman filter on 6D poses to generate smooth and accurate trajectories of object pushing
- Created methods to search and simulate robot actions that imitate experts’ pushing behaviors
Simulated Kitchen Robot Solving Long-horizon Manipulation Tasks with Short Action Library
Feb 2022 – Apr 2022, Carnegie Mellon University, Pittsburgh, PA
- Proposed a two stage optimization method for learning to solve long-horizon manipulation tasks in a simulated kitchen environment in Mujoco
2021
Impedance Control on Delta Robots
Sep 2021 – Dec 2021, Carnegie Mellon University, Pittsburgh, PA
- Implemented an impedance controller on a robot with delta legs to soften impact on free-falling
Adaptive Gripper for Sorting Electronics Waste
Sep 2021 – Dec 2021, Carnegie Mellon University, Pittsburgh, PA
- Developed an adaptive gripping policy on a direct-driven gripper to pick up devices of different types and sizes
- Implemented kinematics, closed-loop position feedback control, and impedance control on the gripper
Robot Arm Manipulation from Hand Motion Commands
Jul 2021 – Sep 2021, Carnegie Mellon University, Pittsburgh, PA
Summer Research Intern, Intelligent Control Lab
- Implemented a control interface between the computer vision program and the robot arm hardware
- Conducted real-time demos and experiments that integrate hand motion learning with robot manipulation
Airborne Object Detection and Tracking Challenge
Jul 2021 – Aug 2021, Carnegie Mellon University, Pittsburgh, PA
Summer Research Intern, AirLab
- Benchmarked state-of-the-art deep learning detection models to select the best ones for the challenge
- Training deep learning models on airborne datasets with MMDetection to detect airplanes and helicopters
Autonomous Chess-Playing Robot Arm
Jan 2021 – Apr 2021, University of Michigan
Computer Science Capstone Design Project
- Implemented a computer vision program to identify and locate the pieces with cameras
- Developed control and path planning strategies to grab and move the pieces with a desktop robot arm
MBot Setpoint and Escape Challenge
Jan 2021 – Apr 2021, University of Michigan
EECS 467 Team Projects
2020
Tik Tok Crab Dance Challenge
Nov 2020, University of Michigan
EECS 367 Individual Project
TechLab at MCity: Innoviz LiDAR Data Collection
Jan 2020 – May 2020, Ann Arbor, MI
Software Integration Engineer, TechLab at MCity Program sponsored by Innoviz Technologies
- Predicted the arrival time of autonomous vehicles and pedestrians with LiDAR sensors
- Integrated the prediction into infrastructure communications to optimize traffic efficiency at intersections
Configurable High-Throughput Puncher
Jan 2020 – May 2020, University of Michigan
Mechanical Engineering Capstone Design Project
- Prototyped a mechanical apparatus for high-throughput packaging a stem-cell-based microfluidic platform
- Performed engineering analyses and validation on various design concepts to slump punching cycle time
2019
Color Ball Detection and Collection
Jan 2019 – Apr 2019, University of Michigan
ME 350 Team Project
2018
Remote-Controlled Vehicle Competition
Sep 2018 – Dec 2018, University of Michigan
ME 250 Team Project