Hi, I’m Yazid Marzuk, a passionate Robotics Researcher and current M.S. student in Robotics at University of Michigan.

My ongoing research focuses on Phyical/Embodied AI. In Professor Ram Vasudevan's Robotics and Optimization for Analysis of Human Motion (ROAHM) LAB, I am working on Risk-Aware Mobile Manipulation on a Fetch Robot in a Gaussian Splat representation of the world. I also work on conducting simulation-to-real (Sim2Real) transfer studies, validating algorithms both in simulation and on the real world applications, By the end of this project we aim to have the Robot successfully deliver our coffee orders to us :)

Apart from all this, I also love working on projects that blur the line between science fiction and reality, whether that’s teaching a robot to pick things faster than me, or getting simulation and hardware to finally agree with each other.

At RightHand Robotics, I developed a Teleoperation Pipeline to control an Industrial Robot(UR10e) with less than 50ms latency. I am currently building towards training a Diffusion Model with demonstration data to achieve autonomous palletization.
At Potential Robotics, I was working on autonomous Underwater Remote Operated Vehicle for Surveillance and Mapping upto 300 meters. I also lead the software team at the company and together we developed ROS based Control System, Digital Twin platform(POT-OS) for managing the vehicle from its Onboard Computer and also designed the Power Supply System.

My long-term goal? To help shape the future of Physical AI, robots with real intelligence, embodied in the physical world, that can learn, adapt, and work alongside us in everyday life.

Outside of robotics? You’ll probably find me Kayaking, Traveling or trying out new Sushi places.