This task is equivalent to existing behaviour of wheels control buttons. When browser window is decreased in size the labels should hide and buttons should turn into arrows.

This issue appears in browser console when browser window is shrunk to mobile size.

Currently, our web page contains such credit notice

LEOBOT©2018 Created by SYSTEMDI 
Developed by Yaryna Demkiv

I suggest to update it to include everyone who works on the project :) Here's how I see it

LeoBot©2018 Created by ROS Ukraine Community

where ROS Ukraine Community links to one of our sites, such as

1. Page with list of LeoBot contributors https://github.com/ros-ukraine/leobot/graphs/contributors
2. ROS Ukraine GitHub page https://github.com/ros-ukraine
3. Facebook page https://www.facebook.com/groups/279366215828155

Visit any local gallery or take open source pictures.
Build simulation of one of the gallerie's or museum's room.
Useful material http://gazebosim.org/tutorials?tut=building_editor
Place new world here https://github.com/ros-ukraine/leobot/tree/kinetic-devel/leobot_gazebo/worlds

Create CAD model of robot in any available application and export it to STL or DAE formats.

Start art gallery simulation command doesn't work
roslaunch leobot_launch simulation.launch world_file:=artgallery.world

Error:
[Err] [SystemPaths.cc:410] File or path does not exist["/home/user/workspace/leobot/base/src/leobot/leobot_gazebo/worlds/artgallery.world.world"]
[Err] [Server.cc:380] Could not open file[/home/user/workspace/leobot/base/src/leobot/leobot_gazebo/worlds/artgallery.world.world]

This tasks consists of such steps:

1. Investigate the VR APIs and estimate if they would work for us better than the current approach
2. If that appears to be the case, implement the most suitable API

Technologies to consider:

1. WebVR API. Resources:

• description
• specification
• support in browsers
• manual on Mozilla Dev site

2. WebXR API which is designed as a replacement and successor of WebVR. Extended reality combines virtual reality, augmented reality and mixed reality (promo).

• What is WebXR API?
• specification
• GitHub spec page
• application from Mozilla

Disadvantages of current implementation:

1. Need to touch the screen and confirm the fullscreen mode.
2. Fullscreen mode in Android works only when vr.html is launched from desktop icon as HTML application.
3. No fullscreen mode in iOS. I tested the current code from B#99_Add_landscape_and_lockorientation branch and could not initialise the VR page under iOS in Chrome, Safari and Firefox. The fullscreen popup appeared on the screen, but when I confirmed it, all browsers remained in windowed mode.

Specifications:

1. Gamepad API
2. Gamepad Extensions API

The site "Can I Use" shows a decent support of this technology betwen browsers, so there's a good chance it will work us.

Use cases:

1. Bluetooth Joystick
2. USB gamepad on Android via OTG cable (connection example)

Use this as an example https://github.com/AndriyPt/ros-navigation-examples/tree/kinetic-devel/moon_2dnav

During our test with @SystemDiagnosticss it gave 7 FPS when we just started Gazebo and 4 FPS when we opened the video stream in VR.

Check MJPEGCANVASJS from http://robotwebtools.org/tools.html#widgets

Create HTML, CSS for design which was created in #4
Please place it in the folder ./docs/design/management
You can use design link from #4 comments

Hello everyone.
I wonder what are the plans for delivering consumer live stream from the robot.

So I was thinking since it is museum robot project the picture quality be better at least HD for the consumer camera. Teleop can be done via SD or lower quality of cause.

• Should the HD camera use ROS transport or just push the stream to RTMP server directly and the main web page just will know the live stream URL?
• Since it is a learning platform we may utilize ROS transport and send live stream via messages. We can use this stream later for NN segmentation and put overlays over paintings etc..

Any plans for the type of camera and implementation of delivering this stream to the user?

Steps to reproduce:

1. Run simulation.launch
2. Run teleop.launch from leobot_control
3. Operate the robot and turn left or right
4. Robot starts swaying left to right with about 1 Hz frequency

Create simulation of the robot with the following features:

• Ability to connect remotely and get video stream from camera
• Teleoperate robot remotely
• Automatic obstacles avoidance

Product implementation milestones

1. Run RViz and watch camera stream plus teleoperate robot from keyboard
2. Run web browser and watch camera stream and teleoperate using buttons on the page
3. Run web browser and watch camera stream and send goals to robot using some easy input (navigation with obstacle avoidance)
4. Use stereo camera and project image into VR headset and use keyboard or joystick to teleoperate robot.

Setup differential drive controller

I suggest to add bindings for such keys

Feel free to post your suggestions re: usability as well as on other questions.

Related task: #52

Readme specifies the following command to launch the gmapping:

roslaunch leobot_launch gmapping.launch

It's necessary to create a gmapping.launch file in leobot_launch package.

Create simple cylindrical base with two differential wheels and one caster wheel.

I'm rather new to GitHub so I wanted to ask for advice on branching and merging of project code.

I've performed such steps:

1. Created a branch F#16_web_server for the task "Create web package #16"
2. Added some preliminary code to this branch (empty package)
3. Created a pull request for this code
4. @AndriyPt merged it into the default branch and closed the pull request. When I open the pull request page on GitHub it suggests to delete this branch but so far it remains in the repository.

Next thing I'm going to add more code to implement this issue.

Could you please help me with such questions:

1. Should I delete the branch F#16_web_server as suggested by GitHub and create a new one with further commits for this task?
2. Should I create another pull request when I finish this task?

Thanks :)
Max

It appears that my installation of Ubuntu in VMWare is working too slow. I'm going to try to install it on a USB stick using these recommendations and will get back with results.

1. Short version: https://askubuntu.com/questions/955258/how-do-i-do-a-full-installation-of-linux-onto-a-usb-flash-drive
2. https://askubuntu.com/questions/16988/how-do-i-install-ubuntu-to-a-usb-key-without-using-startup-disk-creator

Create ROS package which will:

• start web server which can run nodejs
• renders page with text "Hello World"

This issue should be fixed on all pages: main.html and vr.html

In theory, when there's a playing video on the page, browser prevents screen from locking. However we do have video streams on our pages, the screen locking still currently takes place in our case.

Course of action:

1. First of all, investigate how to prevent browser from screen locking using the video streams we have and why this currently does not happen. This is the recommended solution.
2. Only in case if we reasonably would not be able to use the first solution, we can use a third party library for this purpose.

Both issues #51 and #52 (when ready) will initialise the roslib to connect from web browser. This code can be unified so that there's one global "ros" JavaScript variable which can be used for different ROS topics.

Add NAV2DJS to right hand side panel on web interface.
It might be used for goal setting or might be just for visualization of current robot position.
http://robotwebtools.org/tools.html#widgets

Create logo of the robot.
Robot name is LeoBot so it should be something with Lion :)

The web page should occupy full screen size in Chrome, Safari and preferably in Firefox (mobile versions).

iOS Safari and Firefox show 2 additional elements by default: address bar at the top and navigation bar at page bottom.

Joystick will be used for teleoperation.
Mobile browser should read commands from joystick and move robot accordingly.
There should be ability to calibrate joystick on the web page.

It should be wheeled robot
It will contain camera, speakers and microphone.
Also there should be monitor or any other device to visualise remote person.

Optimise the duplicated CSS code in selectors:

#head-control-button-left
#head-control-button-left:active
#head-control-button-left:focus
#head-control-button-center
#head-control-button-center:active
#head-control-button-center:focus
#head-control-button-right
#head-control-button-right:active
#head-control-button-right:focus

Related tasks: #52, #62

I have a task to run render a URDF file using this tutorial: http://wiki.ros.org/urdf/Tutorials/Building%20a%20Visual%20Robot%20Model%20with%20URDF%20from%20Scratch. I faced several issues when trying to accomplish it, so I want to share my solutions with the comunity.

I've tried to use the ROS docker containers as described here: http://wiki.ros.org/docker/Tutorials/Docker.

Here are the steps I've taken:

1. Downloaded the ROS docker image
   docker pull ros:kinetic-robot-xenial
2. Tried to start the container with regular commands docker run -it ros and connect to it in new terminal window using docker exec -it <container_name> bash. In this case, terminal commands were wrapping to the same line and overwrote the beginning of the command, so I couldn't actually see the entire command and the terminal ended up rather unusable for me. As I found out, it happened because of this issue: https://unix.stackexchange.com/a/264236

So I decided to install Terminator within the docker container. This requires some additional steps to make it connect to the host X window environment. Here's my solution.

1. Delete the existing container created from ros image (of you created any) with docker container rm <container_name>

2. Create a new container passing the environment variables and virtual FS (copied from LeoBot):
   docker run -it -e DISPLAY -e LOCAL_USER_ID=$(id -u) -v /tmp/.X11-unix:/tmp/.X11-unix:rw ros
   Triggering printenv in docker container should list the variables you passed.
   During this step, when container is running, you can establish a new connection to it by finding out the container name docker container list and connecting with docker exec -it <container_name> bash.

3. When you try to install any program inside the container (e.g. apt-get install terminator) apt will fail:

Reading package lists... Done
Building dependency tree       
Reading state information... Done
E: Unable to locate package terminator

For some reason this container comes with no preconfigured application repository, so you need to add one apt-get -qq update and then apt-get install terminator.

Then, when you try to start the terminator from container, it will fail with such message:

root@55d1d562bb95:/# terminator
No protocol specified
/usr/lib/python2.7/dist-packages/gtk-2.0/gtk/__init__.py:57: GtkWarning: could not open display
  warnings.warn(str(e), _gtk.Warning)
You need to run terminator in an X environment. Make sure $DISPLAY is properly set

If you passed the parameters when creating this container using docker run, the remaining bit of configuration is to allow connections to X window system on your local machine, i.e. run xhost local:root on your host. After this, it should be possible to launch the terminator from container, even without restarting it.

Other useful commands:

List all docker containers (running and stopped):
docker container list -a
Each time you trigger docker run, it creates a new container, so it's a good idea to:

Delete the container using name or ID from previous command:
docker container rm <container_name>

More info on ROS docker images:
https://store.docker.com/images/ros
https://hub.docker.com/_/ros/

This task applies to head rotation (issue #52) and robot driving buttons (#51). Basically, all buttons on the page.

In order to get effective integration test run on CI servers and apply fix for Gazebo 7 the following changes need to be done

1. Introduce leobot base Docker image with fixed Gazebo version (maybe it will be based on OSRF libgazebo 7 image)
2. Introduce IDE image with all source code and IDEs and Gazebo fixes (inherited from the first one)

The first image will be used for CI servers. second one would be used for Development.

Currently the "control-panel" and "small-buttons" blocks have almost identical markup. I suggest to use the "content" CSS instruction to put labels into buttons and merge these HTML blocks into one. Feel free to leave your comments and considerations :)

Create proof of concept solution for video streaming from one of the robot's camera to web browser.

Creation of the physical robot. Extra functionality:

1. Two way communication
2. Automatic finding of the charging station
3. Web UI
4. Follow predefined path with set of way points. There should be possibility to pause execution e.g. excursion in art museum along Van Gogh works.

When I run catkin_make install from kinetic-devel it fails with such message

CMake Error at leobot/leobot_launch/cmake_install.cmake:51 (file):
  file INSTALL cannot find
  "/home/user/workspace/leobot/base/src/leobot/leobot_launch/launch/display.launch".
Call Stack (most recent call first):
  cmake_install.cmake:124 (include)

Travis build also fails (tested in PR #64 after merging changes from the main branch).

Feel free to assign to me if this issue is actual.

Use this as an example https://github.com/AndriyPt/ros-navigation-examples/blob/kinetic-devel/moon_control/launch/teleop.launch

Create simple design for web site with left hand side menu
It will contain camera output in the middle and navigation arrows in the bottom with Forward, Back, Left and Right directions.