Improving Robotic Exploration via RL Graph Pruning             04/2024 - present
--------------------------------------------------------------------------------

This work aims to improve the efficiency of graph-based robotic exploration
algorithms by designing a novel reinforcement learning agent to prune the graphs
used for path planning. Specifically, the goal is to reduce the amount of
information needed for efficient exploration through this pruning.

To do this, I am building off my last project, in which I designed an RRT-based
robotic exploration algorithm that enables a simulated robot to explore an
unknown environment through a frontier-based method. Once I finish
implementation of the RL agent, I plan to use the algorithm I developed for my
last project along with the simulation environment to perform various
experiments validating the performance of the RL agent.

I am currently pursuing this project under the mentorship of a professor at the
University of Tennessee Knoxville.

<<< Go back