This repository contains autonomous navigation research work conducted at the National University of Singapore (NUS) during Summer 2019. Work was developed in Python for ROS/Gazebo simulation environment. For complete written project report, see AutonomousNavigationBasedOnTurtleBot.pdf
This research project also produced a custom ROS package for four wheeled autonomous vehicles. For more information please see ackermann_nav-ROS.
Main categories are:
- Autonomous Navigation to User Destination
- SLAM
- Multi-Robot SLAM
- Model Predictive Control Path and Speed Planning
User input determine which location robot will navigate to next. SLAM map must be created first.
Experimentation with and evaluation of Google Cartographer.
Implementation of teleop control for mapping a simulation environment using gmapping SLAM.
SLAM readings from multiple robots fused into a single map of the environment.
Example of path planning and speed control for adaptive-horizon MPC. Utilizes on-line computation to dynamically avoid obstacles while navigating towards goal/checkpoint location.