Menu

Computational Robotics

Home / portfolio / Computational Robotics

Computational Robotics Projects

Implementing state of the art algorithms for path planning and robot localization

    These projects are the work from a course in Computational Robotics taken with Professor Kostas Bekris, head of the PRACSYS (Physics-aware Research on Autonomous Computational Systems) Lab at Rutgers University. The projects involve path planning algorithms implementations and simulations using tools like ROS and Gazebo on Ubuntu Linux. The first project deals with path planning of a Turtlebot (essentially a circular-shaped robot) in a 2D environment with pygonal obstacles. The second project deals with sampling-based motion planning using the Probabilistic Roadmap (PRM & PRM*) and Rapidly Exploring Random Tree (RRT) algorithms and simulating the motion of path in a Gazebo environment using the Ackermann model of a vehicle. The third project is an implementation of particle filtering for the localization part of SLAM. The projects can be found on GitHub.

Share

Free-Direction A* Solution Path on 2D Grid-Based Map with Polygonal Obstacles. The Red Nodes correspond to the obstacles, yellow nodes correspond to the visited nodes, and blue nodes correspond to the solution path from given initial to final locations.
Probabilistic Road Map for Object in SE(3) with solution path for initial and final pose
Particle Filter Implementation with noisy controls and range sensor data