PhD Candidate, MIT Mechanical Engineering
I am a PhD candidate in Mechanical Engineering at MIT, advised by Professor Edward Adelson. My research focuses on the intersection of visuo-tactile perception and embodied AI, spanning the full robotics pipeline — from tactile sensor design and hardware engineering to foundation model learning and policy deployment for complex real-world manipulation.
I am particularly interested in overcoming the robotic "data bottleneck" by building scalable, interactive data-collection frameworks with richer sensing modalities. In Summer 2025, I interned at the Robotics and AI Institute, where I worked with Team Capture to design a handheld UMI-style device integrating force-torque and tactile sensors for portable, multimodal manipulation data collection.
I received my B.Eng. from the Tsien Excellence in Engineering Program (TEEP) at Tsinghua University in 2020. I also spent time as an undergraduate researcher in the Xu Group at UC San Diego, working on wearable ultrasound imaging and elastography.
In my spare time, I enjoy weightlifting and playing basketball.
Email / Google Scholar / LinkedIn / GitHub
Transferable Tactile Transformers for Representation Learning Across Diverse Sensors and Tasks
CoRL 2024
[project website] [arxiv] [code]
A tactile representation learned from multi-sensors and multi-tasks, and a tactile dataset containing over 3M tactile images collected from 13 sensors and 11 tasks.
TacLink: A Compact Multi-phalanx Finger with Vision-based Tactile Sensing and Proprioception
ICRA 2024
GelLink has three phalanges and two DOFs, actuated by only one motor and visualized by only one camera. A compact mechanism with a mirror-based tactile sensing system achieves a versatile multi-phalanx design with embedded tactile sensing and accurate proprioception.
GelSight Baby Fin Ray: A Compact, Compliant, Flexible Finger with High-Resolution Tactile Sensing
RoboSoft 2023
Flexible mirrors and high-elongation silicone fluorescent paints incorporated into the GelSight Baby Fin Ray enable grasping through clutter and classification of in-shell nuts.
Stretchable ultrasonic arrays for the three-dimensional mapping of the modulus of deep tissue
Nature Biomedical Engineering 2023
[paper]
A stretchable ultrasonic array for serial non-invasive elastographic measurements of tissues up to 4 cm beneath the skin at 0.5 mm spatial resolution.
A wearable cardiac ultrasound imager
Nature 2023
[paper]
A wearable ultrasonic device for continuous, real-time cardiac function assessment using a deep learning model that automatically extracts left ventricular volume waveforms.
The effect of arterial stiffness on the accuracy of cuff-based blood pressure measurement
Extreme Mechanics Letters 2021
[paper]
A theoretical model studying the impact of arterial wall properties on non-invasive blood pressure measurement, revealing that arteriosclerotic patients' blood pressure may be overestimated.
A 32-channel automatic harp robot with a layered modular framework integrating hardware and software for real-time control and machine learning algorithms for human-machine collaborative music performance.
This section will be updated with hobby projects.