Launch hello_world_demo with V-Rep

[Liiila0205]

Thank you for the plugin! Is there a way to load a .cpp file (based on the youbot API) and simulate that in V-Rep?
With your plugin I can send rostopics to the simulation in V-Rep, and the robot executes them. I can also launch .py files, e.g. by
$rosrun youbot_driver_ros_interface youbot_keyboard_teleop.py
and everything works fine.

But my question is: How can I load .cpp-files based on the KUKA youbot API (http://www.youbot-store.com/youbot-developers/software/libraries/kuka-youbot-api), e.g. the hello_world_demo (https://github.com/youbot/youbot_applications)?

[mfueller]

Hi!

the plugin concentrates on the ROS framework and not on the youBot API.

However, you could write your own youBot API Interface for V-Rep.
Have a look on the repository https://github.com/mfueller/RobotManipulation_WS2013
There you have a quite simple interface for the youBot Arm in V-Rep for c++ - you could use this as a basis your your own implementation.

Unfortunately, the youBot API does not provide "defined interfaces" and implement the stuff directly in the code e.g. YouBotBase.cpp and YouBotManipulator.cpp.
Thus, you would need to define a similar class with the same method definition and switch the #include directive to either load the youBot API or your own API. (just as an idea)

BR,
Matthias

[Liiila0205]

Hi Matthias,

thank you for your fast reply! That is a really good advice! I've just
downloaded your repository and wanted to try the example and take a look
at your interface. But unfortunately when I try to run cmake as you
suggest in your readme-file, I get the following error:

$ cmake /home/lena/catkin_ws/src/RobotManipulation_WS2013
-DVREP_DIR:PATH=/home/lena/v-rep
make: v_repExtRemoteApi_Makefile: No such file or directory
make: *** No rule to make target `v_repExtRemoteApi_Makefile'. Stop.

Do you have any idea what's wrong here?

Best wishes,
Lena

On 02.06.2014 14:28, MFueller wrote:

Hi!

the plugin concentrates on the ROS framework and not on the youBot API.

However, you could write your own youBot API Interface for V-Rep.
Have a look on the repository
https://github.com/mfueller/RobotManipulation_WS2013
There you have a quite simple interface for the youBot Arm in V-Rep
for c++ - you could use this as a basis your your own implementation.

Unfortunately, the youBot API does not provide "defined interfaces"
and implement the stuff directly in the code e.g. YouBotBase.cpp and
YouBotManipulator.cpp.
Thus, you would need to define a similar class with the same method
definition and switch the #include directive to either load the youBot
API or your own API. (just as an idea)

BR,
Matthias

—
Reply to this email directly or view it on GitHub
#2 (comment).

[mfueller]

it seems that v-rep has changed the compilation folders for their remoteAPI. I fixed some path in the CMakeLists and pushed the changed. The library is now compiling. However, I can not check the full example since I do not have the youBot OODL on my computer.

Please check if it is working and let me know if you have any further problems.

[Liiila0205]

thank you for your nice help! now the compiling was successful. But when
I start V-rep, load the scene with the youbot am and try to start the
example by

lena@lena-MacBookPro:~/catkin_ws/src/RobotManipulation_WS2013/build$
./example

I got the error:

terminate called after throwing an instance of
'youbot::FileNotFoundException'
what(): ../config/youbot-ethercat.cfg file not found

do you know what's going on here?

Best regards,
Lena

On 02.06.2014 20:51, MFueller wrote:

it seems that v-rep has changed the compilation folders for their
remoteAPI. I fixed some path in the CMakeLists and pushed the changed.
The library is now compiling. However, I can not check the full
example since I do not have the youBot OODL on my computer.

Please check if it is working and let me know if you have any further
problems.

—
Reply to this email directly or view it on GitHub
#2 (comment).

[mfueller]

found the problem. The YOUBOT_CONFIGURATIONS_DIR was not given as a parameter in the example.cpp - I fixed it and pushed the changes.
If it is complaining about the wrong path, you need to change the YOUBOT_CONFIGURATIONS_DIR the the cmake configuration.

• Matthias

[Liiila0205]

thank you Matthias! I modified the example.cpp as you suggested and
compiled it. But unfortunately I still got the same error:

lena@lena-PC:~/catkin_ws/src/RobotManipulation_WS2013$ sudo ./example
Info: Initializing EtherCAT on eth0 with communication thread
terminate called after throwing an instance of 'std::runtime_error'
what(): No EtherCAT slaves found!

Is it possible that something is wrong in my .cfg files (see attached)?

Best wishes,
Lena

On 06/03/2014 01:20 PM, MFueller wrote:

found the problem. The YOUBOT_CONFIGURATIONS_DIR was not given as a
parameter in the example.cpp - I fixed it and pushed the changes.
If it is complaining about the wrong path, you need to change the
YOUBOT_CONFIGURATIONS_DIR the the cmake configuration.

• Matthias

—
Reply to this email directly or view it on GitHub
#2 (comment).

[EtherCAT]
EthernetDevice = eth0
EtherCATUpdateRate_[usec] = 1000
EtherCATTimeout_[usec] = 500 #timeout value in us for tx frame to return to rx
MaximumNumberOfEtherCATErrors = 100
MailboxTimeout_[usec] = 200
BaseJointControllerName = TMCM-174
BaseJointControllerNameAlternative = TMCM-1632
ManipulatorJointControllerName = TMCM-KR-841
ManipulatorJointControllerNameAlternative = TMCM-1610

[JointTopology]
ManipulatorJoint1 = 0
ManipulatorJoint2 = 1
ManipulatorJoint3 = 2
ManipulatorJoint4 = 3
ManipulatorJoint5 = 4

[Gripper]
EnableGripper = true
DoCalibration = true

[GripperBar1]
BarName = gripper_finger_joint_l
BarSpacingOffset_[meter] = 0.0
MaxTravelDistance_[meter] = 0.0115
MaxEncoderValue = 66000
StallGuard2Threshold = 50
StallGuard2FilterEnable = true

[GripperBar2]
BarName = gripper_finger_joint_r
BarSpacingOffset_[meter] = 0.0
MaxTravelDistance_[meter] = 0.0115
MaxEncoderValue = 66000
StallGuard2Threshold = 50
StallGuard2FilterEnable = true

[Joint_1]
JointName = arm_joint_1
GearRatio_numerator = 1
GearRatio_denominator = 156
EncoderTicksPerRound = 4000
TorqueConstant_[newton_meter_divided_by_ampere] = 0.0335
InverseMovementDirection = true
DoCalibration = true
CalibrationDirection = NEGATIV
CalibrationMaxCurrent_[ampere] = 2.0
UpperLimit_[encoderTicks] = -1000
LowerLimit_[encoderTicks] = -580000
trajectory_controller_P = 20
trajectory_controller_I = 1
trajectory_controller_D = 0
trajectory_controller_I_max = 1000
trajectory_controller_I_min = -1000

[Joint_2]
JointName = arm_joint_2
GearRatio_numerator = 1
GearRatio_denominator = 156
EncoderTicksPerRound = 4000
TorqueConstant_[newton_meter_divided_by_ampere] = 0.0335
InverseMovementDirection = true
DoCalibration = true
CalibrationDirection = NEGATIV
CalibrationMaxCurrent_[ampere] = 0.5
UpperLimit_[encoderTicks] = -1000
LowerLimit_[encoderTicks] = -260000
trajectory_controller_P = 20
trajectory_controller_I = 1
trajectory_controller_D = 0
trajectory_controller_I_max = 1000
trajectory_controller_I_min = -1000

[Joint_3]
JointName = arm_joint_3
GearRatio_numerator = 1
GearRatio_denominator = 100
EncoderTicksPerRound = 4000
TorqueConstant_[newton_meter_divided_by_ampere] = 0.0335
InverseMovementDirection = false
DoCalibration = true
CalibrationDirection = POSITIV
CalibrationMaxCurrent_[ampere] = 1.0
UpperLimit_[encoderTicks] = -1000
LowerLimit_[encoderTicks] = -320000
trajectory_controller_P = 20
trajectory_controller_I = 1
trajectory_controller_D = 0
trajectory_controller_I_max = 1000
trajectory_controller_I_min = -1000

[Joint_4]
JointName = arm_joint_4
GearRatio_numerator = 1
GearRatio_denominator = 71
EncoderTicksPerRound = 4000
TorqueConstant_[newton_meter_divided_by_ampere] = 0.051
InverseMovementDirection = true
DoCalibration = true
CalibrationDirection = NEGATIV
CalibrationMaxCurrent_[ampere] = 0.5
UpperLimit_[encoderTicks] = -1000
LowerLimit_[encoderTicks] = -155000
trajectory_controller_P = 20
trajectory_controller_I = 10
trajectory_controller_D = 0
trajectory_controller_I_max = 1000
trajectory_controller_I_min = -1000

[Joint_5]
JointName = arm_joint_5
GearRatio_numerator = 1
GearRatio_denominator = 71
EncoderTicksPerRound = 4000
TorqueConstant_[newton_meter_divided_by_ampere] = 0.049
InverseMovementDirection = true
DoCalibration = true
CalibrationDirection = NEGATIV
CalibrationMaxCurrent_[ampere] = 0.5
UpperLimit_[encoderTicks] = -5000
LowerLimit_[encoderTicks] = -255000
trajectory_controller_P = 20
trajectory_controller_I = 1
trajectory_controller_D = 0
trajectory_controller_I_max = 1000
trajectory_controller_I_min = -1000

[mfueller]

I would say that you are not connected to a youBot.
The example program controls the simulated arm and real arm simultaneously.
Just remove the youbot oodl code from the example.cpp and it shouldnt complain about the connection problem.

[Liiila0205]

Hi Matthias,

oh I see, now I understand. I didn't consider that your application is
for the real robot, too. I removed the code regarding the real youbot
and now it works fine! Thank you very much! Maybe next time I can help
you, too :).

Best wishes,
Lena

On 06/03/2014 04:13 PM, MFueller wrote:

I would say that you are not connected to a youBot.
The example program controls the simulated arm and real arm
simultaneously.
Just remove the youbot oodl code from the example.cpp and it shouldnt
complain about the connection problem.

—
Reply to this email directly or view it on GitHub
#2 (comment).

[Liiila0205]

Hi Matthias,

thanks again for your nice help last time. At the moment I'm trying to get the forward kinematics for the youBot (based on your RobotManipulation2013-gitrepository) working with KDL (no ROS, only based on the youBotAPI).
After getting a good solution for the forward kinematics, I will try to implement the inverse Kinematics. My goal to move the youbot to a cartestian 3D position with it’s endeffector (inverse kinematics). It is important, that the solution is without
ROS.

Regarding the Forward-Kinematics:
If I pass the axis parameter for the position when the youbot has an unfolded Arm:

KDL::ChainFkSolverPos_recursive
fksolver(chain);

KDL::JntArray
q(chain.getNrOfJoints());

q(0)=2.56;
q(1)=1.04;
q(2)=-2.43;
q(3)=1.73;
q(4)=0;

...I get the following result:

[[ 0.891498, -0.379824, -0.246909;
0.366045, 0.925059, -0.101379;
0.266911, 6.95623e-18, 0.963721]
[0.0229922, 0.00944049, 0.225387]]

But when I look at the simulation in V-Rep, the tooltip is on the position: X=0.0237, y=-0.0280, z= 0.6535. I assume the robot actually has to be on position:

(-0.0430236,-0.00324, 0.2256)

Why is the position in V-Rep different?
Where did you get the values for building the chains (DH Notation) from?
Did the KDL Solver worked in your case?

Best wishes,
Lena

[mfueller]

Hi Lena,

On Wed, Jul 23, 2014 at 3:19 PM, Lena notifications@github.com wrote:

Hi Matthias,

thanks again for your nice help last time. At the moment I'm trying to get
the forward kinematics for the youBot (based on your
RobotManipulation2013-gitrepository) working with KDL (no ROS, only based
on the youBotAPI).
After getting a good solution for the forward kinematics, I will try to
implement the inverse Kinematics. My goal to move the youbot to a
cartestian 3D position with it’s endeffector (inverse kinematics). It is
important, that the solution is without
ROS.

Regarding the Forward-Kinematics:
If I pass the axis parameter for the position when the youbot has an
unfolded Arm:

KDL::ChainFkSolverPos_recursive
fksolver(chain);

KDL::JntArray
q(chain.getNrOfJoints());

q(0)=2.56;
q(1)=1.04;
q(2)=-2.43;
q(3)=1.73;
q(4)=0;

...I get the following result:

[[ 0.891498, -0.379824, -0.246909;
0.366045, 0.925059, -0.101379;
0.266911, 6.95623e-18, 0.963721]
[0.0229922, 0.00944049, 0.225387]]

But when I look at the simulation in V-Rep, the tooltip is on the
position: X=0.0237, y=-0.0280, z= 0.6535. I assume the robot actually has
to be on position:

(-0.0430236,-0.00324, 0.2256)

Why is the position in V-Rep different?

There might be still a difference between the simulated youBot and the real
one. Although I thought that it matches.
I have not time to look into that in detail, but if you find any bugs in
the model, please let me know!

Where did you get the values for building the chains (DH Notation) from?
Did the KDL Solver worked in your case?

Please find attached some peace of source code that one of our student used
to control the robot. It should work on the real robot (as far as I know)
Again, please let me know if it works and I will upload it to the
repository (or send me a patch or pull request)

Regards,
Matthias

Best wishes,
Lena

—
Reply to this email directly or view it on GitHub
#2 (comment)
.

[Liiila0205]

Hi Matthias,

thank you for your mail. I'll take a look at the version of you student.

Did you forget to add the attachements?

Best wishes,
Lena

On 07/24/2014 04:08 PM, MFueller wrote:

Hi Lena,

On Wed, Jul 23, 2014 at 3:19 PM, Lena notifications@github.com wrote:

Hi Matthias,

thanks again for your nice help last time. At the moment I'm trying
to get
the forward kinematics for the youBot (based on your
RobotManipulation2013-gitrepository) working with KDL (no ROS, only
based
on the youBotAPI).
After getting a good solution for the forward kinematics, I will try to
implement the inverse Kinematics. My goal to move the youbot to a
cartestian 3D position with it’s endeffector (inverse kinematics).
It is
important, that the solution is without
ROS.

Regarding the Forward-Kinematics:
If I pass the axis parameter for the position when the youbot has an
unfolded Arm:

KDL::ChainFkSolverPos_recursive
fksolver(chain);

KDL::JntArray
q(chain.getNrOfJoints());

q(0)=2.56;
q(1)=1.04;
q(2)=-2.43;
q(3)=1.73;
q(4)=0;

...I get the following result:

[[ 0.891498, -0.379824, -0.246909;
0.366045, 0.925059, -0.101379;
0.266911, 6.95623e-18, 0.963721]
[0.0229922, 0.00944049, 0.225387]]

But when I look at the simulation in V-Rep, the tooltip is on the
position: X=0.0237, y=-0.0280, z= 0.6535. I assume the robot
actually has
to be on position:

(-0.0430236,-0.00324, 0.2256)

Why is the position in V-Rep different?

There might be still a difference between the simulated youBot and the
real
one. Although I thought that it matches.
I have not time to look into that in detail, but if you find any bugs in
the model, please let me know!

Where did you get the values for building the chains (DH Notation)
from?
Did the KDL Solver worked in your case?

Please find attached some peace of source code that one of our student
used
to control the robot. It should work on the real robot (as far as I know)
Again, please let me know if it works and I will upload it to the
repository (or send me a patch or pull request)

Regards,
Matthias

Best wishes,
Lena

—
Reply to this email directly or view it on GitHub

#2 (comment)
.

—
Reply to this email directly or view it on GitHub
#2 (comment).

[mfueller]

Hi Lena,

On Mon, Jul 28, 2014 at 4:03 PM, Lena notifications@github.com wrote:

Hi Matthias,

thank you for your mail. I'll take a look at the version of you student.

Did you forget to add the attachements?

The attachment was there. Anyway, I added it again.

Regards,
Matthias

Best wishes,
Lena

On 07/24/2014 04:08 PM, MFueller wrote:

Hi Lena,

On Wed, Jul 23, 2014 at 3:19 PM, Lena notifications@github.com wrote:

Hi Matthias,

thanks again for your nice help last time. At the moment I'm trying
to get
the forward kinematics for the youBot (based on your
RobotManipulation2013-gitrepository) working with KDL (no ROS, only
based
on the youBotAPI).
After getting a good solution for the forward kinematics, I will try to
implement the inverse Kinematics. My goal to move the youbot to a
cartestian 3D position with it’s endeffector (inverse kinematics).
It is
important, that the solution is without
ROS.

Regarding the Forward-Kinematics:
If I pass the axis parameter for the position when the youbot has an
unfolded Arm:

KDL::ChainFkSolverPos_recursive
fksolver(chain);

KDL::JntArray
q(chain.getNrOfJoints());

q(0)=2.56;
q(1)=1.04;
q(2)=-2.43;
q(3)=1.73;
q(4)=0;

...I get the following result:

[[ 0.891498, -0.379824, -0.246909;
0.366045, 0.925059, -0.101379;
0.266911, 6.95623e-18, 0.963721]
[0.0229922, 0.00944049, 0.225387]]

But when I look at the simulation in V-Rep, the tooltip is on the
position: X=0.0237, y=-0.0280, z= 0.6535. I assume the robot
actually has
to be on position:

(-0.0430236,-0.00324, 0.2256)

Why is the position in V-Rep different?

There might be still a difference between the simulated youBot and the
real
one. Although I thought that it matches.
I have not time to look into that in detail, but if you find any bugs in
the model, please let me know!

Where did you get the values for building the chains (DH Notation)
from?
Did the KDL Solver worked in your case?

Please find attached some peace of source code that one of our student
used
to control the robot. It should work on the real robot (as far as I know)
Again, please let me know if it works and I will upload it to the
repository (or send me a patch or pull request)

Regards,
Matthias

Best wishes,
Lena

—
Reply to this email directly or view it on GitHub

<
#2 (comment)

.

—
Reply to this email directly or view it on GitHub
<
#2 (comment)
.

—
Reply to this email directly or view it on GitHub
#2 (comment)
.

[Liiila0205]

I'm sorry, but I can't find the attachment :/. Where can I see it?

[mfueller]

it seems that github is removing the attachments.

I have commited it to the github repository
https://github.com/mfueller/RobotManipulation_WS2013/blob/master/src/kdl_kinematic_solver_test.cpp

Regards,
Matthias

On Mon, Jul 28, 2014 at 9:55 PM, Lena notifications@github.com wrote:

I'm sorry, but I can't find the attachment :/. Where can I see it?

—
Reply to this email directly or view it on GitHub
#2 (comment)
.

[Liiila0205]

thank you! I've tested it today and as far as I see the forward kinematic calculations work and fit together with the model in V-REP. Tomorrow I'm going to continue testing and keep you up to date.

Best wishes,
Lena

Am 28.07.2014 um 22:58 schrieb MFueller notifications@github.com:

it seems that github is removing the attachments.

I have commited it to the github repository
https://github.com/mfueller/RobotManipulation_WS2013/blob/master/src/kdl_kinematic_solver_test.cpp

Regards,
Matthias

On Mon, Jul 28, 2014 at 9:55 PM, Lena notifications@github.com wrote:

I'm sorry, but I can't find the attachment :/. Where can I see it?

—
Reply to this email directly or view it on GitHub
#2 (comment)
.

—
Reply to this email directly or view it on GitHub.