4150 Porte De Palmas, unit 11, San Diego, CA92122
Aeolus Robotics Corporation, Ltd., Robotics Software Engineering Intern June 2018 – Aug. 2018
⚫ Transformed Gripper model and Quadruped Robot with joints & links from Solidworks into URDF model.
⚫ Developed the packages on ROS in XML through Docker to plug the URDF model into Gazebo.
⚫ Completed grasping motion of the robot Gripper in Gazebo on ROS to cooperate with SIMULATION team.
⚫ Upgraded the Mass Center Analysis of the Robot in Gazebo on ROS.
Aeolus Robotics Corporation, Ltd., Mechanical Engineer Oct. 2017 – May 2018
⚫ Discussed with leader of SLAM team and MANIPULATION team to design System & Mechanical parts of Robot’s Charging Station with two other functions which satisfied the software spec.
⚫ Integrated module’s placement (critical components: Lidar, IR sensors, cameras…) of new generation Robot.
University of California, San Diego, San Diego, California Sept. 2018 – June 2020 (est.)
M.S. in Electrical and Computer Engineering - Intelligent System, Robotics and Control
⚫ Related Coursework: Robotics Sensing & Estimation (SLAM)(SLAM) (SLAM)(SLAM)(SLAM), Robotics Planning & Learning, Computer Vision, Statistical Learning, Machine Learning in Image Processing, Advanced Data Structure, Linear System
Simultaneous Localization and Mapping Jan. 2019 – Mar. 2019
⚫ Implemented a Particle Filter with IMU, odometry and Lidar data to localize a differential-drive robot by Python.
⚫ Built a 2-D occupancy grid map of the environment and color the map’s floor with the information from Kinect.
⚫ Implemented visual-inertial SLAM using Extended Kalman Filter with IMU and Stereo Camera data.
⚫ Built a 2-D landmarks map that generated the visual features and localized the robot.
Object Detection Jan. 2019
⚫ Built a color model with Gaussian MLE to train a classifier to detect different brightness blue barrels in Python.
⚫ Detected and bounded the position of blue barrels in the images by OpenCV with hand-selected labels.
Pokemon Classifier & Generator Sept. 2018 – Dec. 2018
⚫ Built a dataset with 10000 Pokemon images and classify Pokemons in to different types by CNN with accuracy of 16% and improve the result to 24% by adding Autoencoder with ResNet like skip connection in Pytorch.
⚫ Generated new Pokemon images by GAN with making the decoder in the Autoencoder to be the Generator and using Binary Cross Entropy Loss in the Discriminator.
National Taiwan University, Taipei, Taiwan Sept. 2013 – June 2017
B.S. in Mechanical Engineering
⚫ Concentration: Robotics & Control
⚫ Related Coursework: Introduction to Robotics, Data Structure & Algorithms, Intelligent Control, Scientific Computing, Automatic Control, Microprocessor Controlled System, Digital Control System, System Dynamics
Ukulele Learning to Play Robot Sept. 2016 – Jan. 2017
⚫ Developed a pitch tracker through MATLAB to record and track pitches with audio signal processing algorithms.
⚫ Implemented a zero crossing rate filter to eliminate the noise and disturbance in the environment.
⚫ Integrated hardware system with software algorithms on Arduino to power the robot with self-playing ability.
⚫ Programming: Python, C/C++, MATLAB, Arduino, XML
⚫ Operating System: ROS, Linux, Mac OS, Windows
⚫ Tools: Gazebo, Docker, OpenCV, Pytorch, Embedded systems, AWS, Git, UR5 robotics arm control
⚫ Hardware Skills: Hardware Integration, Solidworks, Simulation, 3D Printing, Milling & Drilling, Lathe
Dream Concept Dance Crew, President June 2013 – Now
⚫ Led the crew to win the 2nd Prize of 2013 MRT Asian Street Dance Competition and perform dozens of shows.
ROS, Python, C++, Hardware, CAD,