I am a Ph.D. researcher working on robot learning, reinforcement learning, and autonomous systems. My work focuses on long-horizon decision-making and learning for real robots, with hands-on experience in ROS 2, simulation, teleoperation, real robot control, and industrial automation.
I am interested in robotics roles where learning methods connect with real hardware, autonomy, and deployment.
Move the cursor to change the target.
A selection of work in robot learning, long-horizon control, ROS 2, and autonomous systems.
Developed a latent-skill planning framework for long-horizon control and evaluated it on D4RL, Adroit, and RoboSuite. The method separates high-level planning from low-level action refinement for multi-stage tasks.
Developed masked skill inference for offline RL when part of the decision context is missing or unreliable.
Built and tested a ROS 2 teleoperation interface in Unreal Engine for SCUTTLE and LIMO mobile robots, with multi-robot control and matching digital-twin environments.
Built a closed-loop pipeline that converted motor-cortex activity into robot motion commands. The system used ROS 2 and ros2_control for joint-space velocity control and was tested on a physical robotic actuator.
I work across learning algorithms, simulation, robotics software, and control.
Offline/online reinforcement learning, imitation learning, actor-critic methods, policy refinement, diffusion models, latent skills, and long-horizon sequence decision-making.
ROS 2, ros2_control, teleoperation, robot feedback loops, multi-robot control, and simulator-to-system integration.
Robot-learning evaluation and policy experimentation in MuJoCo, Isaac Sim, RoboSuite, D4RL/Adroit, and digital-twin environments.
Diffusion models, VAEs, sequence modeling, vision-language and VLA models, LoRA/PEFT, preference-based fine-tuning, and RLHF/GRPO.
Python, C++, Linux, GPU training, CUDA-aware experimentation, reproducible evaluation, and MATLAB/Simulink for control-oriented systems.
PLCs, sensors, instrumentation, electromechanical hardware, troubleshooting, root-cause analysis, machine commissioning, and production deployment.
Selected work most relevant to robot learning, reinforcement learning, and autonomous systems.
I am interested in industrial robotics and robot-learning roles where learning methods connect with real systems, simulation, controls, and deployment.