Projects

Deep RL framework for quadruped navigation over rough terrain via NVIDIA Isaac Sim Sim2Real transfer.

Demonstrates a highly adaptable simulation framework that supports complex terrain navigation, modular AMG/AGV Robot configurations, and the transition from classical control to vision-based perception.

Unlike standard frontier exploration which significantly backtracks, this system uses a 'human-like' exploration strategy to achieve sub-5-minute coverage of complex environments.

Physics-aware RNN model for real-time underwater motion planning under environmental disturbances.

Built a LiDAR-Capable, ultrasonic and IR-based Firefighting Robot AMR/AGV (Autonomous Mobile Robot/ Autonomous Ground Vehicle)

Built and competed in the ASNE PEP25 competition for the first time, representing University of Delaware at a national level.

This project focuses on underwater gesture recognition using image detection. By leveraging neural networks and advanced image processing techniques, the objective was to train a model capable of accurately detecting and classifying underwater gestures.

A binary classifier that distinguishes real from fake news articles.