Hello, I’m Weitung! MIT EECS PhD Student at the Signal Kinetics Group. I received my Bachelor’s and Master's degree in EECS at MIT, and my research interests focus on IoT, AI, and robotics. I love doing hands-on works that can bring my ideas to solve real-world challenges. While in my second year at MIT I teamed up with friends and started a company called Nexuni, Co.. The goal is to develop affordable intelligent automation solutions to help clients manage their businesses and improve quality of life in the cities.
My years at MIT have helped me lay rigorous theoretical foundations in the fields of AI, systems, and robotics. Starting a company with intense R&D works has provided me with tremendous experience in the techniques of rapid prototyping and then integrating these quick creations into the development of various technologies for evaluation. My current PhD research focuses on battery-free localization systems and their potential to enable new forms of perception and autonomy. I will continue to use my knowledge and experience to work on things that impact the world.
You can reach me via weitung@mit.edu or LinkedIn.
Selected Projects
Exploiting Synergies between Augmented Reality and RFIDs for Item Localization and Retrieval
RESEARCH PROJECT. 2023.
* IEEE RFID Best Paper Award *
The system combines Augmented Reality headsets with RFID localization, improving user experience and accuracy, achieving 8.6cm median accuracy and 55% faster item retrieval.
A Robotic System for Fine-Grained RFID Localization using RL-based Synthetic Aperture Radar
RESEARCH PROJECT. 2022.
* Master's Thesis *
The system uses SAR and Reinforcement Learning for efficient RFID tag localization, achieving a mean 3D accuracy of 0.244m, outperforming baselines by 86%.
Carday - Total automated parking management solution for garage and roadside parking
NEXUNI CO. 2022.
Implemented total automated parking management solution including LPR, PGS, Automated roadside parking enforcement, and AI video analysis on more than 20 parking lots / roadside parking sectors in Taiwan.
Autonomous Security Robot
NEXUNI CO. 2021.
This is an ongoing development of facility management robots that aim to reduce the labor need and increase the efficiency of labor-intensive works.
Semiconductor Factory Automation
NEXUNI CO. 2021.
We followed the Industrial 4.0 in the semiconductor factory and provided solutions on factory automation upgrades and high-efficiency IC-test handlers.
Reinforcement Learning based Control for Pouring Liquids
TEAM PROJECT. 2021.
The project explores the dynamics of controlling the simple pouring action to minimize the spillage and the time to finish the pour.
Satellite Storage System Design
TEAM PROJECT. 2020.
This is the final design project of 6.033. The goal is to design a storage system for NASA's satellites to support their their ExtraNet transmission system.
MIT Robotics Science and Systems
TEAM PROJECT. 2020.
Hands on robotic team challenge for MIT 6.141. Robotics: Science and Systems class.
DoDay - Automated Douhua Cafe
NEXUNI CO. 2020.
The goal of DoDay is to demonstrate a turn-key automated store solution to maximize the values of limited human resources.
Line Fresh 2020 - CrowdEye
TEAM PROJECT. 2020.
2nd Place of Line Fresh 2020 Global Hackathon. The project provides a platform to link volunteers with visually impaired people and apply object recognition to help them recognize objects.
Intelligent Parking System
NEXUNI CO. 2020.
Powered by the AI technology, the system can also include features such as Parking Guidance System (PGS), Automated Payment Machine (APM), Online parking fee payment etc.
Autonomous Excavation in Granular Materials with Reinforcement Learning
RESEARCH PROJECT. 2020.
It is about achieving a high-level control of the robot on granular materials by training a deep reinforcement learning network on a mix of real data gathered from the robot, and the simulated data gathered from the partial model of the system.
MIT NEET Lunar-Landing Robotics Challenge
SELF PROJECT. 2019.
Designed and built a robot to complete the MIT NEET Lunar-Landing Robotics Challenge.
AI Scene Segmentation
TEAM PROJECT. 2018.
We examined the effects of adding different skip connections on models for scene segmentation evaluated on MIT ADE20K dataset.
Life at MIT
SELF PROJECT. 2018.
Explore the possibilities of AI creating art. Used deep learning techniques to make computer learn different art styles / movement and apply to the iconic MIT great dome.
Fog Projection Screen
SELF PROJECT. 2017.
A DIY Fog Screen that enables hologram-like projection.
Magnetic Positioning System
RESEARCH PROJECT. 2017.
The research aims to develop a single source magnetic positioning system to allow accurate indoor positioning for mini drones. The research receives many honors, including 1st place from Intel ISEF and Google Science Fair global finalist.
Eco-Switch
SELF PROJECT. 2014.
It is an energy saving device that saves and monitor standby power consumption of house appliances. The project received multiple honors including: Taiwan International Science Fair 1st Place, Google Science Fair Regional Finalist, and Bronze medal in Le Concours Lépine