Robotics Engineering @ University of Michigan · Ann Arbor

I build autonomous robotic systems that perceive, reason, plan, and act in complex real-world environments — spanning underwater robotics, visual navigation, autonomous driving, and multi-robot systems.

My current research interests are centered around spatial intelligence, robot learning, VLM/VLA systems, world models, SLAM, and neuro-symbolic reasoning — particularly how learned representations can be combined with geometry, memory, and structured reasoning for robust long-horizon autonomy.

• Spatial Intelligence & Embodied Navigation — Developing VLM-based navigation and predictive world models for long-horizon ObjectNav, combining visual observations, semantic goals, spatial memory, frontier exploration, and learned decision-making.

• Underwater Perception & Autonomy — Building perception systems for autonomous underwater vehicles using multi-camera imagery, stereo depth, sensor fusion, open-vocabulary detection, 3D reconstruction, and SLAM to detect marine life, geological objects, and other open-set targets in challenging underwater environments.

• Autonomous Shipwreck Inspection — Researching full-stack underwater autonomy for search → detection → circumnavigation → 3D coverage and reconstruction, integrating ROS2, semantic perception, mapping, and motion planning.

• Robot Learning & VLA Navigation — Exploring vision-language-action models, reinforcement learning, predictive representations, and long-horizon reasoning for robots operating in unseen environments.

• Autonomous Driving — Building full-stack autonomy pipelines combining multi-modal perception, sensor fusion, planning, and learned policies in simulation and real-world robotic platforms.

Spatial Intelligence · Vision-Language Models · Vision-Language-Action Models · World Models · Robot Learning · SLAM · 3D Perception · Open-Vocabulary Perception · Navigation & Planning · Multi-Robot Systems · Neuro-Symbolic AI

Languages & Systems Python · C++ · Linux · Git · Docker · ROS 2 · MATLAB

ML & Robot Learning PyTorch · Transformers · Reinforcement Learning · VLMs · VLAs · CLIP · DINO · SAM

Perception & Geometry SLAM · Visual Odometry · Stereo Vision · Multi-View Geometry · Sensor Fusion · 3D Reconstruction · Gaussian Splatting · Open-Vocabulary Detection

Planning & Autonomy ObjectNav · Frontier Exploration · Path Planning · World Models · Spatial Memory · Multi-Robot Navigation

Simulation NVIDIA Isaac Sim · Habitat · CARLA · Gazebo · OceanSim

Scalable Spatial Intelligence Lab — Michigan Robotics Researching spatial intelligence, VLM navigation, predictive world models, and long-horizon robot decision-making.

MOBY Robotics Developing underwater perception and autonomy systems using multi-camera vision, stereo geometry, sensor fusion, 3D reconstruction, and open-set object perception.

University of Michigan Robotics Instructional work in robotics, SLAM, localization, mapping, and autonomous systems.

Building robots that can understand the world, reason about what comes next, and act intelligently within it.