Tutorial (RTM Seminar, Ubuntu, Part 2)

Introduction

This document describes the procedure for creating RT-Components for controlling the Raspberry Pi mouse on the simulator.

/jp/node/6198

Downloading materials

Please download the materials first.

 git clone https://github.com/OpenRTM/RTM_Tutorial

The seminar may be held in an environment where you cannot connect to the Internet. In that case, it is stored in the distributed USB memory.

Simulator

The simulator is developed using a physics engine called Open Dynamics Engine (ODE): http and a drawing library (drawstuff) that comes with ODE. It works if OpenGL works, so it should work in most environments.

You can simulate the following robot Raspberry Pi mouse.

/jp/node/6005

Not only the dynamic calculation and contact response of the Raspberry Pi mouse on the simulator, but also the data of the distance sensor reproduces the values ​​close to those of a real robot.

Raspberry Pi mouse specifications

The Raspberry Pi mouse is an independent two-wheel drive mobile robot sold by RT corp..

/ja/node/6550

RaspberryPi Mouse Specification
CPU Raspberry Pi 2 Model B
Motor Stepping motor ST-42BYG020 x2
Motor Driver SLA7070MRPT x2
Distance sensor Red LED + Photo-transistor(ST-1K3) x4
Monitor red LEDs x4
Buzzer x1
Tact switch x3
Battery LiPo3-cells (11.1V) 1000mAh x1

RT-Component to be developed

  • RobotController Component

An RT-Component to control the robot on the simulator by connecting with the RaspberryPiMouseSimulator Component.

Creating the RobotController Component

An RT-Component which can operate the robot on the simulator with the GUI (slider) and automatically stops when the sensor value exceeds a certain level will be created.

/ja/node/6310

Development process

The development process is as follows.

  • Confirmation of development environment
  • Determine component specifications
  • Creating a source code template with RTC Builder
  • Edit source code
  • Checking the operation of components

Confirmation of development environment

The following development environment is assumed. Setup the development environment on Linux (assuming Ubuntu 18.04 here).

Installation of dependent libraries

 sudo apt-get install gcc g++
 sudo apt-get install libomniorb4-dev omniidl omniorb-nameserver
 sudo apt-get install python-omniorb-omg omniidl-python
 sudo apt-get install cmake
 sudo apt-get install doxygen
 sudo apt-get install openjdk-8-jdk

 # Switch to java8 for Ubuntu 18.04
 sudo update-alternatives --config java

Installation of OpenRTM-aist

For ubuntu 18.04 (64bit)

 # Installation of all the OpenRTM-aist packages
 wget https://raw.githubusercontent.com/OpenRTM/OpenRTM-aist/master/scripts/pkg_install_ubuntu.sh
 bash pkg_install_ubuntu.sh -l all --yes
 
After starting eclipse, RTSystemEditor may not be able to connect to the name server. In that case, add your host name to the localhost line in /etc/ hosts.

 $ hostname
 ubuntu1804 <-- hostname is ubuntu1804
 $ sudo vi /etc/hosts

 127.0.0.1       localhost
 chage this line as follows
 127.0.0.1       localhost ubuntu1804

Installation of CMake

 $ sudo apt-get install cmake cmake-gui

Installation of cmake-gui

 $ sudo apt-get install cmake-qt-gui

Installation of Code::Blocks

Code::Blocks is a integrated development environment which supports C/C++ languages. Please install it with the following commands.

 $ sudo apt-get install codeblocks

If you want to get the latest version, enter the following command.

 $ sudo add-apt-repository ppa:damien-moore/codeblocks-stable
 $ sudo apt-get update
 $ sudo apt-get install codeblocks

Installation of Premake

Required for building ODE.

 $ sudo apt-get install premake4 freeglut3-dev

Build RaspberryPiMouseSimulator component

Build the simulator components manually. Enter the following command.

 $ wget https://raw.githubusercontent.com/OpenRTM/RTM_Tutorial_ROBOMECH2019/master/script/install_raspimouse_simulator.sh
 $ sudo sh install_raspimouse_simulator.sh

The seminar may be held in an environment where you cannot connect to the Internet. In that case, start the script in the distributed USB memory.

 $ sudo sh install_raspimouse_simulator_usb.sh

Component specifications

The RobotController has an outport that outputs the target speed, an import that inputs the sensor value, and configuration parameters that set the target speed and the sensor value to stop.

Name of RT-Component RobotController
InPort
Name of InPort in
Type TimedShortSeq
Description Sensor value
OutPort
Name of OutPort out
Type TimedVelocity2D
Description Target velocity
Configuration
Name of configuration parameter speed_x
Type double
Default value 0.0
Constraints -1.5<x<1.5
Widget slider
Step 0.01
Description The speed of straight-ahead
Configuration
Name of configuration parameter speed_r
Type double
Default value 0.0
Constraints -2.0<x<2.0
Widget slider
Step 0.01
Description The rotational speed
Configuration
Name of configuration parameter stop_d
Type int
Default value 30
Description The sensor value to be stopped the robot

About TimedVelocity2D type

We use the TimedVelocity2D type, which is a data type that stores the moving speed of a moving robot on a two-dimensional plane.

     struct Velocity2D
     {
           /// Velocity along the x axis in metres per second.
           double vx;
           /// Velocity along the y axis in metres per second.
           double vy;
           /// Yaw velocity in radians per second.
           double va;
     };
 
 
     struct TimedVelocity2D
     {
           Time tm;
           Velocity2D data;
     };

This data type can store the velocity vx along the X axis, the velocity vy along the Y axis, and the velocity va around the Z axis.

vxvyva represents the velocity in the robot center coordinate system.


/ja/node/6042

vx is the velocity in the X direction, vy is the velocity in the Y direction, and va is the angular velocity around the Z axis.

For a robot with two wheels mounted on the left and right, such as a Raspberry Pi mouse, vy would be 0 assuming no skidding.

The robot is operated by specifying the straight-ahead speed vx and the rotation speed va.

About distance sensor data

The data of the distance sensor of the Raspberry Pi mouse is designed to output a larger value as the distance to the object gets closer.


rpm14.png

Value from sensor Actual distance[m]
1394 0.01
792 0.02
525 0.03
373 0.04
299 0.05
260 0.06
222 0.07
181 0.08
135 0.09
100 0.10
81 0.15
36 0.20
17 0.25
16 0.30

The simulator reproduces this value and outputs it. The RobotController component implements a process that automatically stops when this value exceeds a certain level.

RobotController component template generation

RTCBuilder is used to generate the template for the RobotController component.

Start RTCBuilder

In Eclipse, the folder where you do various tasks is called a "workspace", and as a general rule, all products are stored under this folder. You can create a workspace in any folder that you can access, but this tutorial assumes the following workspace.

  • /home/<User name>/workspace

First, start Eclipse with the following command.

 $ openrtp

You will first be asked for the location of the workspace, so specify the workspace above.

/ja/node/6058

Then, the following Welcome page will be displayed.


/ja/node/6026
Eclipse の初期起動時の画面

The Welcome page is no longer needed, so click the x button in the upper left to close it.

Click the [Open Perspective] button in the upper right.

/ja/node/6026
Switch perspective

Select RTC Builder to start RTC Builder. The RTCBuilder icon for Hammer and RT appears in the menu bar.

/ja/node/6026
パースペクティブの選択

Create a new project

You need to create a new project in RTCBuilder to create the RobotController component.

Click the Open New RTCBuilder Editor icon in the upper left.

/ja/node/6057
Creating a project for RTCBuilder

Enter the project name (here, RobotController) to be created in the Project Name field and click the [Finish] button.

/ja/node/6310

A project with the specified name will be generated and added to the Package Explorer.

/ja/node/6310

he RTC profile XML (RTC.xml) with default values is automatically generated in the generated project.

Launch RTC Profile Editor

When RTC.xml is generated, the RTCBuilder editor should open as the workspace associated with this project. If it doesn't start, double-click RTC.xml in Package Explorer.

/ja/node/6026

Profile information input and code generation

First, select the Basics tab on the far left and enter the basic information. In addition to the RobotController component specifications (name) you decided earlier, enter the outline, version, etc. Items with red labels are required items. Others can be the default.

  • Module name: RobotController
  • Module description: Any (Robot Controller component)
  • Version: Any (1.0.0)
  • Vendor name: Any
  • Module category: Any (Controller)


/ja/node/6310
Input basic information


Then select the Activity tab and specify the action callback to use.

For RobotController component uses the onActivated(), onDeactivated(), onExecute() callbacks. As shown in the figure below, after clicking onAtcivated in (1), check [ON] with the radio button in (2). Just double-clicking the onActivated label toggles [ON] and [OFF]. Follow the same procedure for onDeactivated and onExecute.


/ja/node/6310
Select activity callbacks


In addition, select the DataPort tab and enter the DataPort information. Enter the following based on the specifications you decided above. The variable name and display position are optional and can be left as they are.


  • InPort Profile:
    • Port name: in
    • Data type: TimedShortSeq

  • OutPort Profile:
    • Port name: out
    • Data type: TimedVelocity2D


/ja/node/6310
Input DataPort information


Then select the Configuration tab and enter the Configuration information based on the specifications you just decided. Constraints and widgets are used to change values in the GUI, such as sliders, spin buttons, and radio buttons, when displaying component configuration parameters in RTSystemEditor.

The straight speed speed_x and the rotation speed speed_r make it easier to operate with the slider.


  • speed_x
    • Name: speed_x
    • Data type: double
    • Default value: 0.0
    • Constraints: -1.5<x<1.5
    • Widget: slider
    • Step: 0.01
  • speed_r
    • Name: speed_r
    • Data type: double
    • Default value: 0.0
    • Constraints: -2.0<x<2.0
    • Widget: slider
    • Step: 0.01
  • stop_d
    • Name: stop_d
    • Data type: int
    • Default value: 30
    • Widget: text


/ja/node/6310
Input configuration parameter information


Next, select the Language / Environment tab and select the programming language. Here, select C++ (language). In addition, defaults etc. are not set for the language and environment, and if you forget to specify it, an error will occur at the time of code generation, so be sure to specify the language.

/ja/node/6310
プログラミング言語の選択


Finally, click the Generate Code button on the Basic tab to generate a template for the component.


/ja/node/6310
雛型の生成(Generate)


The generated code set are generated in the workspace folder specified when eclipse is started. You can see your current workspace in File> Switch Workspace ...

Generating the files needed to build with CMake

The code generated by RTCBuilder includes CMakeLists.txt to generate various files required for building by CMake. By using CMake, Visual Studio project files, solution files, Makefiles, etc. can be automatically generated from CMakeLists.txt.

Operation of CMake (cmake-gui)

Use CMake to configure the build environment. First, start CMake (cmake-gui).

 $ cmake-gui
/ja/node/6311
Launching CMake GUI and specifying target directory

At the top of the window, there are text boxes like the one below, so specify the source code location (where CMakeList.txt is) and the build directory, respectively.

  • Where is the soruce code
  • Where to build the binaries

The source code location is where the RobotController component source was generated and where CMakeList.txt resides. The default is <workspace directory>/RobotController.

This directory is set by dragging and dropping from Explorer to cmake-gui without any manual input.

The build directory is the place to store project files, object files, and binaries for building. The location is arbitrary, but in this case it is recommended that you specify a subdirectory of RobotController with a descriptive name, such as <workspace directory>/RobotController/build.

Where is the soruce code /home/<user name>/RobotController
Where to build the binaries /home/<user name>/RobotController\build

Once specified, click the [Configure] button below. Then a dialog like the one below will be opened, so specify the type of project you want to generate. This time, specify CodeBlocks -- Unix Makefiles. If you don't use Code::Blocks, use Unix Makefiles.

/ja/node/6026
Specify the type of project to be generated

If you don't use cmake-gui, you can generate a Makefiles with the following command.

 $ mkdir build
 $ cd build
 $ cmake .. -G "CodeBlocks - Unix Makefiles"

Click the [Finish] button in the dialog to start Configure. If there are no problems, Configuring done will be output in the log window at the bottom, so click the [Generate] button. If Generating done is displayed, the output of the project file, solution file, etc. is completed.

Note that CMake generates a cache file at the stage of Configure, so if you change the settings or change the environment due to trouble etc., delete the cache with [File]> [Delete Cache] and start over from Configure.

Editing headers and sources

Then double-click RobotController.cbp in the build directory you specified earlier to start Code::Blocks.

Edit the header (include/RobotController/RobotController.h) and the source code (src/RobotController.cpp) respectively. Click RobotController.h, RobotController.cpp from the Project explorer of Code::Blocks to open the edit screen.

/ja/node/6311

Code::Blocks may become unstable in a 64-bit environment. In that case, disabling the code completion plugin may work.

Please select [Plugins]>[Manage plugins...].

/ja/node/6057/

Select [code completion] and click the [Disable] button.

/ja/node/6057/

If it doesn't work, try this procedure.

アクティビティ処理の実装

Implementation of activity processing

In the RobotController component, the configuration parameters (speed_x, speed_y) are operated with sliders and the values are output from the outport (out) as the target speed. The value input from import (in) is stored in the variable, and if the value exceeds a certain value, it stops.

The following figure shows the processing contents of onActivated(), onExecute(), and onDeactivated().

/ja/node/6310/
Overview of activity processing


Editing the header file (RobotController.h)

Declare the variable sensor_data to temporarily store the sensor value.

   private:
      int sensor_data[4];    // A variable that stores the sensor value

Editing the source file (RobotController.cpp)

Implement onActivated(), onDeactivated(), onExecute() as shown below.

 RTC::ReturnCode_t RobotController::onActivated(RTC::UniqueId ec_id)
 {
         // Initialize sensor vaiables
         for (int i = 0; i < 4; i++)
         {
                 sensor_data[i] = 0;
         }
 
         return RTC::RTC_OK;
 }

 RTC::ReturnCode_t RobotController::onDeactivated(RTC::UniqueId ec_id)
 {
         // Stopping the robot
         m_out.data.vx = 0;
         m_out.data.va = 0;
         m_outOut.write();
 
         return RTC::RTC_OK;
 }

 RTC::ReturnCode_t RobotController::onExecute(RTC::UniqueId ec_id)
 {
         // Check if data is arrving
         if (m_inIn.isNew())
         {
             // Read input data from InPort
             m_inIn.read();
             for (int i = 0; i < m_in.data.length(); i++)
             {
                 // Storing input data
                 if (i < 4)
                 {
                     sensor_data[i] = m_in.data[i];
                 }
             }
         }
 
         // Determine if robot is stopped only when moving forward
         if (m_speed_x > 0)
         {
             for (int i = 0; i < 4; i++)
             {
                 // Determine if the sensor value is greater than certain value
                 if (sensor_data[i] > m_stop_d)
                 {
                         //Stop if the sensor value is greater than the set value
                         m_out.data.vx = 0;
                         m_out.data.va = 0;
                         m_outOut.write();
                         return RTC::RTC_OK;
                 }
             }
         }
         // If there is no sensor with a value higher than the set value,
        // operate with the value of the configuration parameter.
         m_out.data.vx = m_speed_x;
         m_out.data.va = m_speed_r;
         m_outOut.write();
           return RTC::RTC_OK;
 }

Building on Code::Blocks

Run build

Please click [Build] button oon the Code::Blocks to build the project.


/ja/node/6311
Run build


Checking the operation of RobotController component

Connect the created RobotController to the simulator component and check the operation.

Start NameService

Start the name service to register the component reference as follows.


 $ rtm-naming

Launch the RobotController component

Please start the RobotController component. Run the RobotControllerComp file in the RobotController/build/src folder.

 $ ./RobotControllerComp

Launching the simulator component

After moved to the directory where the RaspberryPiMouseSimulator component is installed ({the directory where the install_raspimouse_simulator.sh script is run}/RasPiMouseSimulatorRTC/build), launch with the following command.

 $ ./src/RaspberryPiMouseSimulatorComp

Component connection

As shown in the figure below, connect the RobotController component and RaspberryPiMouseSimulator component by using RTSystemEditor.

/ja/node/6310
コンポーネントの接続

Activate component

Click the All Activate icon at the top of RTSystemEditor to activate all components. If activated successfully, the component will be displayed in yellow-green as shown below.


/ja/node/6310
Activate components


Operation check

You can change the configuration from the [Edit] button in the configuration view as shown below.


/ja/node/6311

Check if you can operate the [Raspberry Pi] mouse on the simulator by operating the slider.


/ja/node/6310
コンフィギュレーションパラメーターの変更


Operation check on the actual machine

If you have prepared an actual Raspberry Pi mouse on the seminar, you can check the operation on the actual robot, so if you have time to spare, please try it.

The procedure is as follows.

  • Power on the Raspberry Pi mouse
  • Connect to a Raspberry Pi mouse access point
  • Port connection
  • Component activation

Turn on the power

The Raspberry Pi mouse has two power switches, one is the Raspberry Pi power switch and the other is the motor power switch.


rpm8.png


Turn on the power switch on the inside to start the Raspberry Pi.


rpm9.png


Turning off the power

When turning off the Raspberry Pi, do not turn it off directly from the power switch. Press the button in the middle of the three buttons for a few seconds to start shutting down. The Raspbian shutdown will be completed in about 10 seconds, so turn off the power switch after that.


/ja/node/6042

Connect to an access point

Please refer to the following page for how to connect to the access point.

If the network is switched, the component registration to the name server and the port connection may fail, so close all the name server and components once. If you start it after switching the network, there is no problem, so you do not need to stop it.

To restart the Name Server on the RT System Editor, click the "Start Name Service" button again.

/ja/node/6550

Connecting to the Name-Server

Then add 192.168.11.1 with the [Add Name Server] button in the RTSystemEditor.


tutorial_raspimouse0.png tutorial_raspimouse1.png



Then you will see the following two RTCs.

/ja/node/6311

RaspberryPiMouseRTC is an RT-Component for controlling Raspberry Mouse, which is being developed in the Robot System Design Laboratory of Meijo University.

Port connection

Connect the RaspberryPiMouseRTC and RobotController components in the RTSystemEditor as follows.

/ja/node/6310

Turn on the motor

Please turn on the power switch of the motor before operation. Be sure to turn off the power of the motor frequently.


rpm10.png


Activation

Then activate the RTC and you will be able to operate the Raspberry Pi mouse.

Download

latest Releases : 2.0.0-RELESE

2.0.0-RELESE Download page

Number of Projects

Choreonoid

Motion editor/Dynamics simulator

OpenHRP3

Dynamics simulator

OpenRTP

Integrated Development Platform

AIST RTC collection

RT-Components collection by AIST

TORK

Tokyo Opensource Robotics Association

DAQ-Middleware

Middleware for DAQ (Data Aquisition) by KEK