Robotic-Autonomy-Project

MAE-6592 final class project

Install dependencies:

sudo apt update
sudo apt-get install ros-melodic-teleop-twist-keyboard ros-melodic-gazebo-ros-pkgs ros-melodic-gazebo-ros-control

ROS Depth Camera in Gazebo

See http://gazebosim.org/tutorials/?tut=ros_depth_camera for full installation tutorial.

Install ROS nodes

To install the ROS nodes clone this repo into your catkin workspace:

cd /home/$USER/catkin_ws/src/
git clone https://github.com/km5es/Robotic-Autonomy-Project.git
cd ..
catkin_make

NOTE: In case it is not already so, make the human_detector node executable by:

chmod +x src/Robotic-Autonomy-Project/human_detector/scripts/human_detector.py

Run "human" model in Gazebo:

roslaunch kbot_description kbot_base_rviz_gazebo.launch

Run teleop:

roslaunch kbot_simple_control kbot_control_teleop.launch

Optional: Set Fixed Frame to 'camera_link' and add a PointCloud2 and Image with Topics '/camera/depth/points' and '/camera/color/image_raw' respectively in the open Rviz window.

Run human_detector node

rosrun human_detector human_detector.py

The human truth data will be saved in /home/$USER/track_human.csv
The robot truth data will be saved in /home/$USER/track_robot.csv
EKF data will be dumped in /home/$USER/track_EKF.csv
The data folder contains some example data and a video.

NOTE

If the robot gets stuck in a wall there is an explicit teleop launch to get it out:

roslaunch kbot_simple_control follower_control_teleop.launch

km5es/Robotic-Autonomy-Project