Autonomous Driving Stack on a Real GEM e4
Built and validated an end-to-end autonomous driving stack integrating perception, behavioral planning, local planning, and closed-loop control on a real GEM e4 platform.
Robotics · Autonomy · Perception · Machine Learning
I'm an ECE graduate student at UIUC building intelligent systems across perception, state estimation, planning, control, and machine learning — from simulation to real-world deployment.
I'm a graduate student in Electrical & Computer Engineering at UIUC working across robotics, autonomy, perception, and machine learning. The through-line is the same: build the stack that turns sensors into action — calibration, 3D perception, state estimation, planning, and closed-loop control — and measure it where it actually has to work.
That has meant a full autonomy stack on a real GEM e4, vision-based navigation on a quadrotor, 3D localization on KITTI, and a vision-guided manipulator — plus two years building a healthcare guidance robot in an intelligent-control lab. This past summer at Texas Instruments I worked the other end of the problem: production ML on 338M+ manufacturing records, with backtesting, experiment tracking, and deployment.
From autonomous vehicles and aerial robots to 3D perception, manipulation, and production ML.
Built and validated an end-to-end autonomous driving stack integrating perception, behavioral planning, local planning, and closed-loop control on a real GEM e4 platform.
Developed a vision-based autonomous navigation pipeline combining monocular perception, pose estimation, trajectory optimization, and feedback control for waypoint navigation.
LiDAR-camera calibration · Graph-SLAM · 3D detection · Multi-object tracking
Built components of a 3D autonomy perception stack including sensor calibration, point-cloud processing, object detection, localization, and tracking.
Vision-guided 6-DoF UR3 arm: camera calibration, object detection, inverse kinematics, and closed-loop pick-and-place.
Built an interactive healthcare guidance robot using Whisper, retrieval-augmented generation, and a mobile service-robot interface.
LLM-assisted diffusion framework that turns editing instructions into executable multimodal plans for multi-object image edits.
Co-first author. Robot system integrating navigation, motion guidance, and hospital service workflows on the Temi platform.
Co-first author. Speech-to-text, retrieval-augmented generation, and an LLM, reaching 93% recommendation accuracy in deployment testing, later 99.6% with a reselect model.
Perception, navigation, and closed-loop robot behavior in one Isaac Sim scene — synthetic camera and LiDAR into detection, into planning, into control on the simulated robot. Progress and clips land here as it comes together.
Open to full-time robotics, autonomy, perception, and machine learning roles starting after I graduate in May 2027.