I'm Wells, a former national champion in Electrical Installations. After conquering the electrical field, I'm now expanding my horizons into the realm of information technology.
2018
~2022
2014
~2017
Jul 2022
~Present
Feb 2021
~Jul 2022
Jan 2019
~Present
Jun 2020
~Jan 2021
Aug 2021
~Apr 2022
Jan 2018
~Sep 2019
HTML | |
CSS | |
Javascript |
Rails | |
Bash | |
PostgreSQL |
Sketchup | |
PLC | |
KNX |
Nginx | |
Git | |
Docker |
Hochschule Esslingen Course Selection Simulator is a simulation course selection system developed specifically for Hochschule Esslingen. This system aims to assist students in resolving confusion during the course selection process by providing features such as course selection simulation, course search, and course information. It is implemented using Rails and React.
NTUST Senior is a LINE chatbot designed specifically for new non-departmental students at National Taiwan University of Science and Technology (NTUST). Developed by myself to address the confusion faced by non-departmental freshmen every year, this project provides features such as course selection assistance, graduation requirements, and campus announcements. It is implemented using Rails.
Sketchup is a commonly used 3D modeling software in the architecture industry. For this project, the requirement is to automatically deploy reinforcement within the structural elements. In traditional cases, modelers have to manually calculate and construct sample reinforcement based on the dimensions, layers, and types of reinforcement required for the structural elements. They then repeatedly copy the sample reinforcement to specific locations, which is a repetitive and time-consuming task. With this plugin, however, by inputting parameters, the plugin can directly generate reinforcement that matches the structural elements. It is implemented using Ruby.
This project is a reconnaissance robot designed for exploring indoor spaces and surveying partially collapsed buildings. To adapt to complex terrains, we chose a hexapod robot as the base, which offers greater advantages in overcoming terrain obstacles despite its slower movement speed compared to wheeled robots. The project utilizes reinforcement learning techniques to simulate various terrains on a computer, enabling the robot to automatically adjust its movement patterns based on the terrain. When used in disaster relief scenarios, the robot provides valuable insights into the interior of collapsed buildings. We have equipped the robot with an optical radar to collect data and generate a point cloud representation of the entire space. Additionally, for easier location of trapped individuals, the robot is equipped with a camera that utilizes YOLOv4 for real-time image recognition, facilitating search and rescue operations.