ridgeback / ridgeback

Common packages for Ridgeback.

ridgeback

Common packages for Ridgeback, including messages and robot description. These are packages relevant to all Ridgeback workspaces, whether simulation, desktop, or on the robot's own headless PC.

ridgeback's Issues

Undefined substitution argument namespace

On the kinetic-devel branch, trying to load the robot description using

roslaunch ridgeback_description description.launch

leads to the following error:

Undefined substitution argument namespace
when processing file: /path/to/catkin_ws/src/ridgeback/ridgeback_description/urdf/ridgeback.gazebo
included from: /path/to/catkin_ws/src/ridgeback/ridgeback_description/urdf/ridgeback.urdf.xacro

It might be due to changes from b47792b, which added a reference to a namespace argument, but does not seem to define it anywhere.

URDF with ros force based move problem

In your description file, I have noticed that PublishOdomTF was set to "false" refer to the plugin named ros_force_based move. Then you use Odometry from IMU instead if I understand clearly.
My question is

1. What is the problem with PublishOdomTF, so you stop publishing this message?
2. I have tried to modify Ridgeback description and launch file to make multiple robot simulation. However, there are too many parameters to modify from EKF localization node and others. Could you please suggest me how to make multiple Ridgeback simulation?

And I also have tried the original ros force based move plugin from hector_gazebo. It has the problem that I can't get the robot position in y axis from Rviz. Do you have the same problem with me ?

Shared pointers for melodic

Starting in melodic, the urdf parser uses std::shared_ptr instead of boost::shared_ptr. I've made the fixes in a branch on my fork: jonbinney@9fcc361

I'm not PRing these into kinetic-devel of this repo since they are breaking changes. Could either create a melodic-devel branch or I could add a bunch of preprocessor stuff to change the code depending on distro. Preference?

Ridgeback description (urdf)

Hi,

I am planning to make use of this urdf file in order to simulate the ridgeback in v-rep.

As I looked through the urdf file, it does not seem dynamically reasonable.
Since the front/rear rocker is stuck in axle link, the robot won't move even the wheels rotate.
Do I have to actuate other joints together in order to move the robot?

Additionally, am I right that the robot description is only representing the non-holonomic robot?
Is it possible to represent the omnidirectional movement of the robot?

It would be grateful if you could provide any explanation with the files.

Many thanks in advance!

ROS Noetic?

Hi,

Is there a plan to support ROS Noetic for the Ridgeback for Gazebo?

Thanks

Ridgeback urdf does not move in Pybullet simulation engine

Hi. I used the Ridgeback urdf in Pybullet Gym engine. But I found that it can not move using any command to control the wheels. But i check the urdf file. The wheel is continuous. Why in the simulation the joint value is always zero. Really confused. Could you give me some idea on that?

Many Thanks for that!

melodic-devel branch removed?

PR #24 went into the branch melodic-devel with some breaking changes, but I no longer see that branch.

There were some quick fixes to remove deprecation warnings in melodic.

[ WARN] : global_costmap: Pre-Hydro parameter "static_map" unused since "plugins" is provided
[ WARN] : global_costmap: Pre-Hydro parameter "map_type" unused since "plugins" is provided
[ INFO] : global_costmap: Using plugin "static_layer"
[ INFO] : global_costmap: Using plugin "inflater_layer"

[ WARN] : local_costmap: Pre-Hydro parameter "static_map" unused since "plugins" is provided
[ WARN] : local_costmap: Pre-Hydro parameter "map_type" unused since "plugins" is provided
[ INFO] : local_costmap: Using plugin "obstacles_layer"
[ INFO] : local_costmap: Using plugin "inflater_layer"

[ WARN] : Parameter pdist_scale is deprecated. Please use the name path_distance_bias instead.
[ WARN] : Parameter gdist_scale is deprecated. Please use the name goal_distance_bias instead.

Ridgeback Moving in RViz but not Gazebo

Hi,
I've been following this tutorial here: https://docs.clearpathrobotics.com/docs/ros1noetic/robots/indoor_robots/ridgeback/tutorials_ridgeback#simulating-ridgeback.

When I try to run: roslaunch ridgeback_gazebo ridgeback_world.launch and I keep getting a message that the urdf spawner exits.

[urdf_spawner-11] process has finished cleanly
log file: /home/sinarasi/.ros/log/3b010fac-32f9-11ee-96ed-1df6afe6d206/urdf_spawner-11*.log

The log file mentioned above doesn't appear to help with debugging:
[rospy.client][INFO] 2023-08-04 15:00:52,145: init_node, name[/urdf_spawner], pid[50395]
[xmlrpc][INFO] 2023-08-04 15:00:52,145: XML-RPC server binding to 0.0.0.0:0
[xmlrpc][INFO] 2023-08-04 15:00:52,145: Started XML-RPC server [http://sinarasi:33507/]
[rospy.impl.masterslave][INFO] 2023-08-04 15:00:52,146: _ready: http://sinarasi:33507/
[rospy.registration][INFO] 2023-08-04 15:00:52,146: Registering with master node http://localhost:11311
[rospy.init][INFO] 2023-08-04 15:00:52,146: ROS Slave URI: [http://sinarasi:33507/]
[xmlrpc][INFO] 2023-08-04 15:00:52,146: xml rpc node: starting XML-RPC server
[rospy.init][INFO] 2023-08-04 15:00:52,146: registered with master
[rospy.rosout][INFO] 2023-08-04 15:00:52,146: initializing /rosout core topic
[rospy.rosout][INFO] 2023-08-04 15:00:52,148: connected to core topic /rosout
[rospy.simtime][INFO] 2023-08-04 15:00:52,148: initializing /clock core topic
[rospy.simtime][INFO] 2023-08-04 15:00:52,149: connected to core topic /clock
[rosout][INFO] 2023-08-04 15:00:52,150: Loading model XML from ros parameter robot_description
[rosout][INFO] 2023-08-04 15:00:52,153: Waiting for service /gazebo/spawn_urdf_model
[rospy.internal][INFO] 2023-08-04 15:00:52,396: topic[/rosout] adding connection to [/rosout], count 0
[rospy.internal][INFO] 2023-08-04 15:00:52,701: topic[/clock] adding connection to [http://sinarasi:38241/], count 0
[rosout][INFO] 2023-08-04 15:00:52,757: Calling service /gazebo/spawn_urdf_model
[rosout][INFO] 2023-08-04 15:00:52,941: Spawn status: SpawnModel: Successfully spawned entity
[rospy.core][INFO] 2023-08-04 15:00:52,942: signal_shutdown [atexit]
[rospy.internal][INFO] 2023-08-04 15:00:52,943: topic[/rosout] removing connection to /rosout
[rospy.internal][INFO] 2023-08-04 15:00:52,943: topic[/clock] removing connection to http://sinarasi:38241/
[rospy.impl.masterslave][INFO] 2023-08-04 15:00:52,943: atexit
~
This exiting is problematic since when I launch RViz with roslaunch ridgeback_viz view_robot.launch and control it with the interactive controller, the ridgeback only moves in RViz and not Gazebo.

Any guidance would be appreciated.

Ridgeback does not move in simulation

I'm picking up this question again as I dont fully understand how moving the ridgeback does work under the hood in Gazebo.

To my experimental knowledge the ridgeback does not move at all when the force-based plugin is removed (the wheels do spin due to the MecanumDriveController, however), but it seems not (enough) force is applied - if any. Enabling contact points in Gazebo reveals that all 4 wheel do have ground contact indeed. I cant move in any direction using the demo (melodic-devel) and steering via keyboard (or rviz marker server). So is the force plugin a 'cheat' in order to make it work in Gazebos physics engine (ODE in my case) or is there something I have overlooked? This should be documented clearly to avoid confusions for the users

Thank you very much in advance :)!

Originally posted by @uahic in #29 (comment)

Issues when simulation is reset

Hi there!

As soon as the simulation is reset using the service /gazebo/reset_simulation, Ridgeback stops publishing the odometry topic. The problem seems to be in the gazebo plugin libridgeback_ros_force_based_move.so. I assume that the problem occurs due to a variable in this plugin which stores the last timestep when a topic was published and which is not reset when the simulation is reset, but I don't have insights into this pre-compiled stuff. Could this be fixed soon? For reinforcement learning, the support of the reset_simulation service is crucial.

Thanks!
Julien

Ridgeback collision failed

I added the collision in urdf, but the robot arm can also go through the top and chassis link. Could you show us how to fix that?

Many thanks for that

Can not subscribe any message from topic "/ridgeback_velocity_controller/odom"

I get some problem with odom and tf:

(1) I can not subscribe any message from topic "/ridgeback_velocity_controller/odom", But the "/status" topic is working properly. I can get the output voltage information of the motor driver from the "/status" topic. In addition, I can also control robot movement through the "/cmd_vel" topic.

(2) Because no message is available from the topic "/ridgeback_velocity_controller/odom", "/odometry/filtered" topic is also problematic. "/odometry/filtered" topic has twist information, its pose information is constantly increasing, even though my robot is still. Eventually the transformation of odom to base_link is also wrong.
screen_shot_1

(3) When I executed "roslaunch ridgeback_viz view_robot.launch", the rviz interface told me that: the tf information is incomplete and there are multiple links that are not connected to the odom. In addition, there are some problems with time in the tf transformation.

No transform from [front_left_wheel_link] to [odom] 
No transform from [front_right_wheel_link] to [odom] 
No transform from [front_rocker_link] to [odom] 
No transform from [rear_left_wheel_link] to [odom] 
No transform from [rear_right_wheel_link] to [odom]

For frame [front_laser]: No transform to fixed frame [odom]. TF error: [Lookup would require extrapolation into the future. Requested time 1545108932.845590104 but the latest data is at time 36280661.233775787, when looking up transform from frame [front_laser] to frame [odom]]

Message removed because it is too old (frame=[imu_link], stamp=[34.113000000])

screen_shot_2
screen_shot_3
screen_shot_4

wheel_separation_x, wheel_separation_y parameters don`t work

I have found a bug with wheel_separation parameters.
You should take out these code block from outer "if" operator to make it work.

ridgeback/ridgeback_control/src/mecanum_drive_controller.cpp

Lines 416 to 424 in 69b9a89

	else
	{
	ROS_INFO_STREAM("Wheel seperation in X: " << wheel_separation_x_);
	ROS_INFO_STREAM("Wheel seperation in Y: " << wheel_separation_y_);
	// The seperation is the total distance between the wheels in X and Y.
	wheels_k_ = (wheel_separation_x_ + wheel_separation_y_) / 2.0;
	}