AI · Robotics · Intelligent Hardware
I work on intelligent hardware: soft sensors, the electronics that read them, and the machine learning that helps turn their signals into a sense of touch.
Ph.D. Candidate in Mechanical Engineering · Data Sciences Institute Doctoral Fellow · University of Toronto
I'm a Ph.D. candidate in Mechanical Engineering at the University of Toronto and a Data Sciences Institute (DSI) Doctoral Fellow.
My research focuses on soft tactile-sensing skins, and most of my work has been on the hardware behind them: fabricating the sensors, designing the electronics and circuit boards that read them, writing embedded firmware, and applying machine learning to the signals they produce. I'm especially interested in edge AI that runs directly on the sensing hardware, and in how this kind of sensing can support embodied AI.
Thesis: A Biomimetic Tactile-sensing Skin for Prosthetic Digits. GPA 4.0/4.0.
Thesis: Elderly Fall Detection by Smart Wearable Devices. Ranked 1st in class.
Soft sensors, the electronics that read them, and the machine learning that interprets their signals.
The focus of my Ph.D. Soft, biomimetic sensing skins that can register contact, shape, and texture, together with the data-acquisition and signal-processing pipeline that reads their multichannel EIT measurements, and machine-learning models designed with embedded hardware in mind.
A miniaturized, wearable development board for electrical impedance tomography (EIT) robotic skins, small enough to sit right next to a sensor patch. I designed the measurement electronics, schematic and PCB layout, and PlatformIO firmware, and released it all as open-source hardware. Published at IEEE-RAS Humanoids 2024.
Custom-shaped soft sensing skins made with 3D printing, along with the fabrication process behind them. Presented at the 10th World Congress of Biomechanics (WCB 2026) in Vancouver.
More than 150 multi-layer PCBs designed in Altium, from power electronics to wearable devices, along with FPGA- and microcontroller-based controllers for AC/DC actuators and the 20+ robots they have driven.
Fall detection for older adults using wearable inertial sensors and convolutional neural networks, including new ways to turn motion signals into images for training (Distinguished Paper Award, ICBME 2021), and a commercialized VR system that visualizes force-plate and motion data.
The tools I use to go from a circuit board to a trained model.
Claude Code for research, firmware development, robotics programming, and policy design.
TabPFN for small-data tabular prediction; Jev for calibrated, structured decisions.
Deep learning (CNNs), classical ML (KNN, K-means), signal processing for EIT and IMU data, computer vision.
150+ multi-layer PCBs in Altium Designer; FPGA and microcontroller controllers (ARM, AVR, Arduino); analog front-ends.
Python, C/C++, MATLAB, PlatformIO, Git.
3D printing of soft hydrogel–elastomer structures, SolidWorks, COMSOL.
Press coverage of my research.
“The complexity of the project allows us to break it down into smaller pieces for teams with different expertise.”
Peer-reviewed papers and conference presentations.
I'm always happy to talk about AI for hardware, embodied AI, tactile sensing, or collaboration opportunities.