Autonomous-Mobile-Robotics

This repository contains from the scratch ROS implementation of robot autonomy algorithms.

System requirements

1. Ubuntu 20.04 LTS
2. ROS Noetic

---

Installation steps

• sudo apt-get install ros-noetic-hector*
• git clone [email protected]:sManohar201/Autonomous-Mobile-Robotics.git
• cd Autonomous-Mobile-Robotics
• catkin_make or catkin build
• source devel/setup.bash

---

List of topics covered in the course

1. Perception
2. Sensor Fusion
3. Localization
4. Simultaneous Localization and Mapping
5. Path Planning

---

TODO: Perception

• combine front and rear laser into a single source of laser data.
• feature extraction algorithm, line, blobs, and corners for now.
• data association for slam and localization.
• Observation model

TODO: Sensor Fusion

• base template for sensor fusion
• Implement EKF
• Implement UKF
• Compare performance between sensor fusion package and robot_pose_ekf package
• odometry to trajcetory package to compare how odom estimates drift from true state values.

TODO: Localization

• Iterative closest point (should I move this to perception)
• Kalman filters with iterative solution.
• Monte Carlo Localization.

TODO: SLAM

• Occupancy grid mapping (half done improve bresenham rasterisation algorithm)
• EKF slam with unknown correspondencies.

TODO: Path Planning