Skip to content

Latest commit

 

History

History
151 lines (99 loc) · 6.17 KB

README.md

File metadata and controls

151 lines (99 loc) · 6.17 KB
Raw

Running the demos under Robotics SDK Docker based ROS environment

This sections covers the details on setting up the ROS environment for running the available demos under this repository.

Setting Up Robotics SDK Docker Container Environment

The first step towards running the demos under ROS is to setup the Docker environment.

Setting up J7 SK Board

Please follow the instructions in the Robotics SDK User Guide Documentation on Docker Setup for ROS1 to work with ROS1 on J7 SK board.

Once the robotics SDK docker container is set up, follow the instructions below to build a docker container for running the demos:

root@tda4vm-sk:/opt/robot# edgeai-robotics-demos/docker/build.sh

If the above commands are successful, you should have a docker image built successfully. An example output based on 8.2 release is shown below.

root@tda4vm-sk:/opt/robot# docker images

REPOSITORY              TAG                 IMAGE ID            CREATED             SIZE
j7-ros-noetic_scuttle   8.2                 a7b4a5e3fcdb        About an hour ago   3.25GB
j7-ros2-foxy            8.2                 50c5af3ef3c2        5 days ago          1.8GB
j7-ros-noetic           8.2                 71bb852e17e8        5 days ago          3.15GB
arm64v8/ubuntu          20.04               b5e8d9885aa9        6 days ago          65.6MB

Setting up UBUNTU PC

Please follow the instructions in the Robotics SDK User Guide Documentation on Set Up Docker Environment on the Remote PC for Visualization.

Once the robotics SDK docker container is setup on PC, follow the instructions below to clone the edgeai-robotics-demos repository and SCUTTLE_ROS:

user@pc:~/j7ros_home$ cd ros_ws/src
user@pc:~/j7ros_home/ros_ws/src$ git clone https://github.com/TexasInstruments/edgeai-robotics-demos.git
user@pc:~/j7ros_home/ros_ws/src$ git clone -b noetic https://github.com/scuttlerobot/SCUTTLE_ROS/ scuttle_ws

Then build a docker container for running the demos.

user@pc:~/j7ros_home/ros_ws/src$ edgeai-robotics-demos/docker/build_pc.sh

If the above commands are successful then you should have a docker image built successfully. An example output based on 8.2 release is shown below.

user@pc:~/j7ros_home/ros_ws/src$ docker images

REPOSITORY                    TAG       IMAGE ID       CREATED             SIZE
pc-ros-noetic_scuttle         8.2       f60c6c407314   About an hour ago   3.22GB
pc-ros-noetic                 8.2       4dfd0b15c047   3 days ago          3.19GB
ubuntu                        20.04     ba6acccedd29   3 days ago          72.8MB

Building the demos

Log on to the Docker container on the SK board by running the following command.

root@tda4vm-sk:/opt/robot# edgeai-robotics-demos/docker/run.sh

Follow the commands below to build the demos.

root@j7-docker:~/j7ros_home/ros_ws$ mkdir -p ros1_build
root@j7-docker:~/j7ros_home/ros_ws$ cd ros1_build
root@j7-docker:~/j7ros_home/ros_ws/ros1_build$ catkin_make --source /opt/robot
root@j7-docker:~/j7ros_home/ros_ws/ros1_build$ source devel/setup.bash

Running the demos/applications

The following out-of-box demo applications are available under ROS1.

OpenCV based ball follower

The details of this demo could be found under OpenCV based ball follower.

[J7 SK BOARD] To launch the openCV ball follower demo, run the following:

root@j7-docker:~/j7ros_home/ros_ws/ros1_build$ source devel/setup.bash
root@j7-docker:~/j7ros_home/ros_ws/ros1_build$ roslaunch opencv_subject_follower opencv_follower.launch

[Visualization on Ubuntu PC] Run the following commands from within the ROS1 docker for visualizing the output.

user@pc:~/j7ros_home$ ros_ws/src/edgeai-robotics-demos/docker/run_pc.sh
root@pc-docker:~/j7ros_home/ros_ws$ rviz -d /root/j7ros_home/ros_ws/src/edgeai-robotics-demos/ros1/rviz/follower.rviz

Note that PC_IP_ADDR and J7_IP_ADDR should be set before running Docker container on the ubuntu PC. Update these variables in setup_env_pc.sh and source it by the following command:

user@pc:~/j7ros_home$ source setup_env_pc.sh

EdgeAI based ball follower

The details of this demo could be found under EdgeAI based ball follower.

[J7 SK BOARD] To launch the openCV ball follower demo, run the following:

root@j7-docker:~/j7ros_home/ros_ws/ros1_build$ source devel/setup.bash
root@j7-docker:~/j7ros_home/ros_ws/ros1_build$ roslaunch edgeai_subject_follower edgeai_follower.launch

[Visualization on Ubuntu PC] Run the following commands from within the ROS1 docker for visualizing the output.

user@pc:~/j7ros_home$ ros_ws/src/edgeai-robotics-demos/docker/run_pc.sh
root@pc-docker:~/j7ros_home/ros_ws$ rviz -d /root/j7ros_home/ros_ws/src/edgeai-robotics-demos/ros1/rviz/follower.rviz

Lidar based navigation

The details of this demo could be found under Lidar based Navigation.

[J7 SK BOARD] To launch the Lidar based navigation demo, run the following:

root@j7-docker:~/j7ros_home/ros_ws/ros1_build$ source devel/setup.bash
root@j7-docker:~/j7ros_home/ros_ws/ros1_build$ roslaunch lidar_navigation lidar_navigation.launch

[Visualization on Ubuntu PC] Run the following commands from within the ROS1 docker for visualizing the output.

user@pc:~/j7ros_home$ ros_ws/src/edgeai-robotics-demos/docker/run_pc.sh
root@pc-docker:~/j7ros_home/ros_ws$ rviz -d /root/j7ros_home/ros_ws/src/edgeai-robotics-demos/ros1/rviz/lidar_navigation.rviz

Scuttlebot simulator on Gazebo

The details of this demo could be found under scuttlebot_simulator package.

Documentation

Documentation on the SCUTTLE ROS library can be found at the following link: