Home Service Robot
November 2020
Embedded Files
ROS navigation stack using Djikstra's algorithm to find path through occupancy grid.
Gazebo environment
Loop closures found while performing Real Time Appearance Based Mapping (RTAB)
3D point cloud of environment created by RTAB map
Check out this project's source code!
In the fall of 2020, I completed Udacity's Software Engineering Nanodegree. Five course projects culminate in this capstone.
The goal of the project is to navigate to virtual objects, pick them up, and drop them off on a known occupancy map previously generated with the Graph SLAM algorithm (above, left). Robot Operating System (ROS) nodes written in C++ communicate with one another via a publish and subscribe architecture, providing the robot's odometry measurement and laser scan sensor data as input to an Adaptive Monte Carlo Localization (AMCL) algorithm, or particle filter. With the estimated pose and map in hand, Djikstra's algorithm is used to generate a path to the goal pose.
This project exists entirely in software, and has inspired me to apply the concepts in a current hardware robotics project - coming soon!
Page updated
Google Sites
Report abuse