NIST and UMD are affiliated by PREP program which allows UMD faculty to work at NIST. https://prep.umd.edu/

Developed ROS 2-based frameworks for real-time testing and validation of perception models on both physical autonomous vehicles and CARLA simulations.

Debugged sensor data from cameras, lidar, and GPS in vehicle to extract key features for sensor fusion, supporting accurate object detection, localization, and perception in autonomous vehicle scenarios.

Designed and executed complex experimental setups for sensor-driven perception scenarios, including multi-modal input from vehicle sensors.

Implemented a Level 1 autonomous traffic sign recognition feature using ROS 2 and Python, demonstrating end-to-end perception and classification capabilities.

Pioneered the design of a novel cutting-edge framework methodology for testing autonomous vehicle (AV) systems interaction.

Developed path-planning algorithm to generate optimum trajectories for a neurosurgical procedure using DFS.

Generated point cloud from MRI data to create cost map and reward map for critical brain structures.

Enhanced code base to improve maintainability and reduce time required to run software by 80%.

Created a user-friendly GUI for neurosurgeons to generate surgical trajectories in MATLAB.

Collaborated with neurosurgeons to compare generated trajectories in Slicer visualization tool.

Enhanced the existing testing pipeline to deploy new URDF files to simulation 40% faster.

Created a unified robotics description format file (URDF) of a complex forklift for use in simulation.

Designed and implemented test scenarios to verify dynamic stability and inform the robot’s motion planning strategy.

Analyzed camera coverage of onboard 3D depth cameras to ensure situational awareness using simulation.