Ackermann Robot Bring-Up & Native Autonomy
Brought up the vehicle controller, validated safety-gated motion, and built the sensing, odometry, and mapping foundation for autonomy.
Read case study →NYU M.S. Robotics · AI4CE Researcher
Robotics systems from firmware to autonomy.
I am a robotics systems engineer who builds and debugs complete robotic systems across embedded control, state estimation, and robot learning—from STM32 bring-up and CAN-based actuation to visual-inertial SLAM and language-conditioned policies.
Seeking robotics software, embedded robotics, autonomy, controls, and robot-learning internship opportunities.
0.031 mbest reported SLAM trajectory error after optimization
3,630EuRoC frames processed in ORB frame-to-map evaluation
33 Hzembedded drivetrain telemetry logging rate
30diffusion-policy evaluation rollouts with failure analysis
How I Work
Evidence before claims
I enjoy robotics problems that cross boundaries between firmware, sensors, controls, estimation, and learned behavior. My work emphasizes safe hardware bring-up, explicit system architecture, reproducible experiments, and honest separation between measured results and planned capability.
I am currently completing an M.S. in Robotics at NYU and contributing to CREO / AI4CE research on mobile manipulation and vision-language-action systems.
Selected Work
Systems engineering in evidence
Three projects that best show hardware ownership, algorithmic depth, and end-to-end validation.
Stereo Visual-Inertial SLAM
Built a full EuRoC pipeline from stereo tracking through bundle adjustment, loop closure, and sparse mapping.
Read case study →
Automatic Transmission Controller
Integrated a physical 24 V drivetrain with concurrent sensing, encoder-tracked shifting, data logging, and efficiency analysis.
Read case study →Current Experience
Research at NYU CREO / AI4CE
Student Researcher
CREO / AI4CE Lab · NYU
May 2026–Present
- Support multi-camera teleoperation dataset pipelines and Linux hardware/firmware diagnostics.
- Patched a PD + feedforward-torque command path, correcting torque encoding, mode activation, and CAN frame structure for safe manipulator actuation.
- Contribute to research on vision-language-action and vision-language models for a dual-arm mobile manipulator.
Background
Education and direction
New York University
M.S. Robotics, Mechatronics Track
Expected May 2027
I am targeting robotics systems, embedded robotics, autonomy, controls, and robot-learning roles where I can connect algorithms to reliable hardware behavior.