Computer Science

In this project, I made a robot dog for STEM education based on 3D printing and Arduino development kit. On top of this, I completed the integration of ESP32 and MPU6050, which gave the robot the ability of motion detection. Through the calculation of the motion data and the implemented PID control system, I completed a very good Self-Balance function. In addition, I also tried to apply the self-balance algorithm to the motion state which improves the motion posture and optimize the performance of the robot dog walking on certain terrain. Finally, I sorted out the problems encountered in the development process and some important principles into a document for STEM education, which can help others better understand and apply related content.

AI versus human players in games has been an increasingly popular topic, especially after the victory of AlphaGo. This project aims at developing a Connect Four AI for STEM education that utilizes a mobile application and a robotic arm to play with human opponents. In order to accomplish the objective, multiple software and hardware tools and methods were employed to design the system workflow of the Android application. Furthermore, experiments were conducted to select the best approaches. Based on the results, computer vision with OpenCV circle and color detection was used to recognize the board, and an optimized minimax algorithm with Alpha-Beta pruning was implemented to calculate the next best move. With the completed product, players have an advanced Connect Four gaming experience by playing against the perfect AI, which can be utilized in STEM education by demonstrating the strength of AI in making decisions and recognizing objects. However, with the limitations in the application and robotic arm, the product can further be improved to enhance usability and gaming experience in the future.

In this project, the goal is to develop the lost and found functionality of the patrol robot supplied by Airport Authority Hong Kong, where the robot patrols around designated areas and the camera of the robot will capture the surroundings. With object detection performed with an AI model, the robot will find lost items in the checkpoints and then store information of 2nd class objects, such as wallets, into a database. If lost belongings are found, the robot will report to the staff at the Hong Kong international Airport through a Telegram report channel and update a web portal which shows records of lost items.

Having trouble finding groupmates for your course project or just want to find a study group? iCollab has got you covered! Use the app to find enthusiastic groupmates to work together! No more hassle of trying to find groupmates in an unfamiliar classroom!

Have you ever lost any items in HKU? Or found any lost items but didn’t know what to do?
HKUSearchU helps HKUers to search and report lost items.

You can view all found lost items in a list or a map. Each record contains the item type and description, found time and location, retrieving method and an image.

If you found something, use the app to fill in a simple form so other users can view the record.

Our Motivation is to minimize the waste of food due to passing the expiry date and better stock management. Our Main Features include Expiry Date Tracking, Record Product ,Information,Notification before expiry dates and Share with family and groups. Other Features include Other Remark memo (e.g. Price) and Show Calories / Nutrition Facts.

With a rapidly ageing population in Hong Kong, there is a need to provide more volunteer services to the elderly. We introduce HINCare, an app that encourages mutual-help in the community. HINCare adopts the concept of timebanking, where each user has his/her own account to store the amount of time (”time credits”). Users earn time credits by providing services to others. The user can use the time credits to purchase other services. Moreover, HINCare uses HIN (Heterogeneous Information Network) to recommend helpers to elders.

This project provides a new way to measure incompatibility between quantum measurements, with notable advantages such as its ease of computation as well as it being a metric for one-dimensional quantum measurements – something that enables clustering quantum measurements using conventional machine learning techniques such as K-Medoids clustering.