tumftm / carla-autoware-bridge Goto Github PK

Thank you so much for providing this awesome bridge between Carla and Autoware!

I am currently trying to study the implication of using Carla's synchronized mode based on this documentation.

Q1: Is there a specific configuration of Carla that we must use (e.g., synchronous mode with fixed time-step) for this bridge to work properly?

Looks like by default, Carla runs in asynchronous mode, and I assume that is the default option preferred by this bridge.

Q2: Do you see any potential issue when running Carla in synchronous mode with this bridge? What can go wrong or should I expect no difference?

Thank you for your time and consideration!

Is it possible to run your setup in Singularity? I want to run it on an HPC cluster which doesn't allow Docker.

Thanks for this work.

I get some problem following the tutorial in readme. Any help will be appreicated.

1. After starting carla0.9.15(localhost) and the package(docker), starting autoware(localhost), the rviz2 panel open and close frequently. I catch the message " cannot receive pointcloud" in the output log when starting autoware.

2. So i intent to view the pointcloud and image topic, only Qos setting Best Effort and fixed frame lidar_front can make two topic data visualable.
   

I realized that the sensor data and tf topic published is dismatch in frame.
sensor frame (as expect):

• camera_front_link
• lidar_front
• lidar_rear

tf topic(may be abnormal):

I want know the following things:
a. First, the dismatch is normal? If not , is there a way to fix the dismatch?
b. Is it right that autoware can not work as expect due to the frame dismatch?

my enviroment:

• memory :64G GPU:nvidia RTX3090 processor:inter i9-11900K
• ubuntu22.04
• ros-hunble
• cuda--11.7
• cudnn-8.9.5.30
• tensort-8.6.0.12
• autoware tag: v1.0

Finally I attach the output in terminal when starting autoware.
autoware_carla.log

I am not using the carla-autoware bridge yet. I am using carla as an external source to get data for autoware. I have only one lidar attached to vehicle. ndt_scan_matcher is being used to do localization.

I keep getting the message:

Have you ever faced this before. If so, how can I fix it?

• Fix to docker pull tumgeka/carla-autoware-bridge:latency (not latest)
• ...

Hi, I'm working on an application that requires an autoware 2024.01 docker image. Before, I could download it by using: docker pull ghcr.io/autowarefoundation/autoware-universe:humble-2024.01-cuda-amd64, however, since last Friday I'm unable to. Docker shows the following error:

I never needed to login before to use this command or to download this image, also non-Autoware images do work properly, like the default hello-world image used to test docker itself.

The command provided in the bridge's README file no longer works for me. Could it be that the 2024.01 image has been removed by Autoware? Is there any other compatible image I can use instead?

Thanks for your help.

Do you have any other lanelet2 maps, my project need other maps?
I saw that you used software for conversion, and I also saw that the paper mentioned the need for manual modification. May I ask if you have any experience and methods in modifying maps?

Currently, manual control is not functioning out of the box. We need to explore whether there is a viable solution for integrating it into our system. A solution has already been implemented here: main...dilloncsr:Carla-Autoware-Bridge:main

The Traffic Light Detection should at least work for Town10HD, either using the original map-based approach or integrating the new Autoware Fine Detection

Autoware expects the vehicle_state_report message. This essentially only says that the vehicle will or will not execute a control message in the current state. We should integrate this message into our bridge so that the warning no longer appears

Thanks for giving such an excellent project.
I just have a simple problem that when I use "2D Pose Estimate" button in autoware to set the Car's position, why the Car in Carla changed first, and then in autoware?
I just echo two topics "/initialpose" and "/sensing/gnss/pose_with_covariance", and find that Carla publishes the Gnss sensor msg continuously, but the "/initialpose" is published only one time when I use the "2D Pose Estimate" button.

And How Carla get the autoware' s "/initialpose" topic msg to change the Position in Carla, where is the code about it?

Hi, I reproduced step by step the tutorial found on youtube. After running carla, bridge and starting autoware, I see correctly the vehicle placed in autoware and in the RecognitionResultOnImage box there is the correct image. After selecting the path and starting the simulation, the vehicle moves correctly in both autoware and carla but the image in RecognitionResultOnImage remains stuck at the initial frame. Is this a problem with the camera? Do you have a solution to the problem? Thanks in advance

Error Information:
...
[system_error_monitor-22] [ERROR] [1717456182.973000596] [system.system_error_monitor]: vehicle_state_report msg is timeout...
[component_container_mt-76] [INFO] [1717456184.590174946] [default_ad_api.node.operation_mode]: perception component state is unhealthy. Autonomous is not available.
[component_container_mt-70] [WARN] [1717456189.312649806] [planning.scenario_planning.lane_driving.behavior_planning.behavior_velocity_planner.traffic_light_module]: Failed to calculate stop point and insert index
[component_container_mt-70] [WARN] [1717456189.312701596] [planning.scenario_planning.lane_driving.behavior_planning.behavior_velocity_planner.traffic_light_module]: the traffic signal data associated with regulatory element id is not received
[component_container_mt-76] [INFO] [1717456190.375389685] [default_ad_api.node.operation_mode]: perception component state is unhealthy. Autonomous is not available.
[system_error_monitor-22] [ERROR] [1717456192.562001896] [system.system_error_monitor]: vehicle_state_report msg is timeout...
[component_container_mt-76] [INFO] [1717456196.004920461] [default_ad_api.node.operation_mode]: perception component state is unhealthy. Autonomous is not available.
[component_container_mt-70] [WARN] [1717456198.794032538] [planning.scenario_planning.lane_driving.behavior_planning.behavior_velocity_planner.traffic_light_module]: Failed to calculate stop point and insert index
[component_container_mt-70] [WARN] [1717456198.794093723] [planning.scenario_planning.lane_driving.behavior_planning.behavior_velocity_planner.traffic_light_module]: the traffic signal data associated with regulatory element id is not received
[component_container_mt-76] [INFO] [1717456201.759390581] [default_ad_api.node.operation_mode]: perception component state is unhealthy. Autonomous is not available.
[system_error_monitor-22] [ERROR] [1717456202.042719076] [system.system_error_monitor]: vehicle_state_report msg is timeout...
[component_container_mt-76] [INFO] [1717456207.499882634] [default_ad_api.node.operation_mode]: perception component state is unhealthy. Autonomous is not available.
[component_container_mt-70] [WARN] [1717456208.344776105] [planning.scenario_planning.lane_driving.behavior_planning.behavior_velocity_planner.traffic_light_module]: Failed to calculate stop point and insert index
[component_container_mt-70] [WARN] [1717456208.344850717] [planning.scenario_planning.lane_driving.behavior_planning.behavior_velocity_planner.traffic_light_module]: the traffic signal data associated with regulatory element id is not received
[system_error_monitor-22] [ERROR] [1717456211.634117040] [system.system_error_monitor]: vehicle_state_report msg is timeout...
[component_container_mt-76] [INFO] [1717456213.284798141] [default_ad_api.node.operation_mode]: perception component state is unhealthy. Autonomous is not available.
...

Autoware underwent significant control changes after Version 2024.01. These modifications have resulted in strong braking during curves. This issue should be thoroughly investigated and optimized. Additionally, we plan to integrate a vehicle interface for our Carla vehicle in the long term.

Currently it is not clear that the frame_id represents the ros2 topic name. Perhaps we should explain this somewhere

Dear developers,

I have successfully built Carla, carla-ros-bridge, and Autoware for Ubuntu on a local desktop.
I could not find any documentation for making a local build of Carla-Autoware-Bridge for the system.
Where can I find the right information or do you have any direct solution for me?

thank you very much in advance

yours

Hi,I tried to fill in view:=true and there was no image.It is not clear whether this is related to the error in the red box?

Hello, I'm currently working with Carla and Autoware using this Bridge, and I'm encountering really severe performance issues. Autoware seems almost frozen and scripts like the generate_traffic.py always crash due to timeouts. After 2.5 hours, the simulation only triggered 171 ticks (50 ms). Here's some considerations to take into account:

Computer Specs

• Ubuntu 22.04
• 64 GB of Memory
• Intel Core i9-9980XE CPU @ 3.00 GHz with 36 cores
• 2 Nvidia Quadro P4000 GPUs

Carla

• Carla is making one of my CPU cores work always at 100%. I believe Carla is the responsible of this because the problem persists even when I shut down everything else, and resolves once I shut down Carla.
  
• I'm currently using options like -RenderOffScreen and -quality-level=Low for performance.

Autoware

• Because of the project I'm working on, I need to record plenty of data from the Carla-Autoware simulations. I'm using ros2bag to record around 30 Autoware topics.
• I'm using a custom sensor kit based on the default bridge's sensor kit. In this case, I removed the rear lidar, but added two RGB 2k Cameras, with their respective Tensorrt_yolo processes.
• Both Tensorrt_yolo processes are running in the second GPU, while the rest of processes run in the fist one, according to nvidia-smi.

These performance issues seem completely anomalous. Is there any performance option I may have missed? I'm following the commands shown in the main README file to start everything, but maybe some configuration options in Carla or Autoware are preventing the platform from properly using the GPU or consume more resources than needed.

Hello developers,

First of all, thank you for your efforts. I am a newbie in ROS, so I followed your video step by step to build and run the project. When I run
ros2 launch carla_autoware_bridge carla_aw_bridge.launch.py port:=1403 town:=Town10HD

I recieved this error:

root@sebla:/tum# ros2 launch carla_autoware_bridge carla_aw_bridge.launch.py port:=1403 town:=Town10HD [INFO] [launch]: All log files can be found below /root/.ros/log/2024-05-07-13-14-34-932809-sebla-418 [INFO] [launch]: Default logging verbosity is set to INFO [INFO] [bridge-1]: process started with pid [419] [INFO] [carla_spawn_objects-3]: process started with pid [423] [INFO] [set_initial_pose-4]: process started with pid [425] [INFO] [carla_autoware_bridge-2]: process started with pid [421] [INFO] [carla_ackermann_control_node-5]: process started with pid [427] [bridge-1] [INFO] [1715087675.573591676] [carla_ros_bridge]: Trying to connect to localhost:1403 [bridge-1] [WARN] [1715087675.579357745] [carla_ros_bridge]: Version mismatch detected: You are trying to connect to a simulator that might be incompatible with this API. Client API version: 0.9.13-20.04-1-gb428f20-dirty. Simulator API version: 0.9.15 [bridge-1] WARNING: Version mismatch detected: You are trying to connect to a simulator that might be incompatible with this API [bridge-1] WARNING: Client API version = 0.9.13-20.04-1-gb428f20-dirty [bridge-1] WARNING: Simulator API version = 0.9.15 [bridge-1] [INFO] [1715087677.420100570] [carla_ros_bridge]: Trying to create a Traffic Manager server in 8000 with synchronous mode set to: True [bridge-1] [INFO] [1715087679.636941157] [carla_ros_bridge]: Loading town 'Town10HD' (previous: 'Carla/Maps/Town10HD_Opt'). [bridge-1] terminate called after throwing an instance of 'carla::client::TimeoutException' [bridge-1] what(): time-out of 5000ms while waiting for the simulator, make sure the simulator is ready and connected to localhost:1403 [ERROR] [bridge-1]: process has died [pid 419, exit code -6, cmd '/tum/install/carla_ros_bridge/lib/carla_ros_bridge/bridge --ros-args -r __node:=carla_ros_bridge --params-file /tmp/launch_params_smeomh10 --params-file /tmp/launch_params_kgfsefg8 --params-file /tmp/launch_params_nxb9mhp0 --params-file /tmp/launch_params_4vvsa8dn --params-file /tmp/launch_params_yv0zrvxg --params-file /tmp/launch_params_g9nleb8h --params-file /tmp/launch_params_rwqjxcyr --params-file /tmp/launch_params_4wpc8vcz --params-file /tmp/launch_params_76sryu98 --params-file /tmp/launch_params_ndc29za4 --params-file /tmp/launch_params_mjkopeh5 -r /carla/ego_vehicle/sensing/camera/traffic_light/image:=/sensing/camera/traffic_light/image_raw -r /carla/ego_vehicle/sensing/camera/traffic_light/camera_info:=/sensing/camera/traffic_light/camera_info -r /carla/ego_vehicle/sensor/lidar/front:=/sensor/lidar/front -r /carla/ego_vehicle/sensor/lidar/rear:=/sensor/lidar/rear -r /carla/ego_vehicle/sensor/gnss:=/sensor/gnss -r /carla/ego_vehicle/sensor/imu:=/sensor/imu']. [INFO] [launch]: process[bridge-1] was required: shutting down launched system [INFO] [carla_ackermann_control_node-5]: sending signal 'SIGINT' to process[carla_ackermann_control_node-5] [INFO] [set_initial_pose-4]: sending signal 'SIGINT' to process[set_initial_pose-4] [INFO] [carla_spawn_objects-3]: sending signal 'SIGINT' to process[carla_spawn_objects-3] [INFO] [carla_autoware_bridge-2]: sending signal 'SIGINT' to process[carla_autoware_bridge-2] [carla_ackermann_control_node-5] Traceback (most recent call last): [carla_ackermann_control_node-5] File "/tum/install/carla_ackermann_control/lib/carla_ackermann_control/carla_ackermann_control_node", line 33, in <module> [carla_ackermann_control_node-5] sys.exit(load_entry_point('carla-ackermann-control', 'console_scripts', 'carla_ackermann_control_node')()) [carla_ackermann_control_node-5] File "/tum/build/carla_ackermann_control/src/carla_ackermann_control/carla_ackermann_control_node.py", line 595, in main [carla_ackermann_control_node-5] roscomp.shutdown() [carla_ackermann_control_node-5] File "/tum/build/ros_compatibility/src/ros_compatibility/__init__.py", line 84, in shutdown [carla_ackermann_control_node-5] rclpy.shutdown() [carla_ackermann_control_node-5] File "/opt/ros/humble/local/lib/python3.10/dist-packages/rclpy/__init__.py", line 126, in shutdown [carla_ackermann_control_node-5] _shutdown(context=context) [carla_ackermann_control_node-5] File "/opt/ros/humble/local/lib/python3.10/dist-packages/rclpy/utilities.py", line 58, in shutdown [carla_ackermann_control_node-5] return context.shutdown() [carla_ackermann_control_node-5] File "/opt/ros/humble/local/lib/python3.10/dist-packages/rclpy/context.py", line 102, in shutdown [carla_ackermann_control_node-5] self.__context.shutdown() [carla_ackermann_control_node-5] rclpy._rclpy_pybind11.RCLError: failed to shutdown: rcl_shutdown already called on the given context, at ./src/rcl/init.c:241 [carla_autoware_bridge-2] Traceback (most recent call last): [carla_autoware_bridge-2] File "/tum/install/carla_autoware_bridge/lib/carla_autoware_bridge/carla_autoware_bridge", line 33, in <module> [carla_autoware_bridge-2] sys.exit(load_entry_point('carla-autoware-bridge', 'console_scripts', 'carla_autoware_bridge')()) [carla_autoware_bridge-2] File "/tum/build/carla_autoware_bridge/carla_autoware_bridge/carla_autoware_bridge.py", line 24, in main [carla_autoware_bridge-2] rclpy.spin(carla_autoware_bridge) [carla_autoware_bridge-2] File "/opt/ros/humble/local/lib/python3.10/dist-packages/rclpy/__init__.py", line 222, in spin [carla_autoware_bridge-2] executor.spin_once() [carla_autoware_bridge-2] File "/opt/ros/humble/local/lib/python3.10/dist-packages/rclpy/executors.py", line 739, in spin_once [carla_autoware_bridge-2] self._spin_once_impl(timeout_sec) [carla_autoware_bridge-2] File "/opt/ros/humble/local/lib/python3.10/dist-packages/rclpy/executors.py", line 728, in _spin_once_impl [carla_autoware_bridge-2] handler, entity, node = self.wait_for_ready_callbacks(timeout_sec=timeout_sec) [carla_autoware_bridge-2] File "/opt/ros/humble/local/lib/python3.10/dist-packages/rclpy/executors.py", line 711, in wait_for_ready_callbacks [carla_autoware_bridge-2] return next(self._cb_iter) [carla_autoware_bridge-2] File "/opt/ros/humble/local/lib/python3.10/dist-packages/rclpy/executors.py", line 608, in _wait_for_ready_callbacks [carla_autoware_bridge-2] wait_set.wait(timeout_nsec) [carla_autoware_bridge-2] KeyboardInterrupt [set_initial_pose-4] [INFO] [1715087685.701706057] [default]: Cancelled by user. [set_initial_pose-4] Traceback (most recent call last): [set_initial_pose-4] File "/tum/install/carla_spawn_objects/lib/carla_spawn_objects/set_initial_pose", line 33, in <module> [set_initial_pose-4] sys.exit(load_entry_point('carla-spawn-objects', 'console_scripts', 'set_initial_pose')()) [set_initial_pose-4] File "/tum/build/carla_spawn_objects/src/carla_spawn_objects/set_initial_pose.py", line 65, in main [set_initial_pose-4] roscomp.shutdown() [set_initial_pose-4] File "/tum/build/ros_compatibility/src/ros_compatibility/__init__.py", line 84, in shutdown [set_initial_pose-4] rclpy.shutdown() [set_initial_pose-4] File "/opt/ros/humble/local/lib/python3.10/dist-packages/rclpy/__init__.py", line 126, in shutdown [set_initial_pose-4] _shutdown(context=context) [set_initial_pose-4] File "/opt/ros/humble/local/lib/python3.10/dist-packages/rclpy/utilities.py", line 58, in shutdown [set_initial_pose-4] return context.shutdown() [set_initial_pose-4] File "/opt/ros/humble/local/lib/python3.10/dist-packages/rclpy/context.py", line 102, in shutdown [set_initial_pose-4] self.__context.shutdown() [set_initial_pose-4] rclpy._rclpy_pybind11.RCLError: failed to shutdown: rcl_shutdown already called on the given context, at ./src/rcl/init.c:241 [carla_spawn_objects-3] [ERROR] [1715087685.703310463] [default]: Could not initialize CarlaSpawnObjects. Shutting down. [ERROR] [carla_ackermann_control_node-5]: process has died [pid 427, exit code 1, cmd '/tum/install/carla_ackermann_control/lib/carla_ackermann_control/carla_ackermann_control_node --ros-args -r __node:=carla_ackermann_control --params-file /tum/install/carla_ackermann_control/share/carla_ackermann_control/settings.yaml --params-file /tmp/launch_params_9xp6y8_5']. [ERROR] [carla_autoware_bridge-2]: process has died [pid 421, exit code -2, cmd '/tum/install/carla_autoware_bridge/lib/carla_autoware_bridge/carla_autoware_bridge --ros-args -r __node:=carla_autoware_bridge --params-file /tmp/launch_params_pj91b36j --params-file /tmp/launch_params_yzzk7lph']. [INFO] [launch]: process[carla_autoware_bridge-2] was required: shutting down launched system [ERROR] [set_initial_pose-4]: process has died [pid 425, exit code 1, cmd '/tum/install/carla_spawn_objects/lib/carla_spawn_objects/set_initial_pose --ros-args -r __node:=set_initial_pose --params-file /tmp/launch_params_vbmvu47x --params-file /tmp/launch_params_sfva7zj2']. [INFO] [carla_spawn_objects-3]: process has finished cleanly [pid 423]

CARLA is running and it doesnt give any errors. What can I do for it?

Thank you in advance.

I am currently struggling to generate a steer_map.csv which is used by raw_vehicle_cmd_converter node. I am trying the learning based vehicle calibrator (https://github.com/pixmoving-moveit/learning_based_vehicle_calibration). I would like to know how did you create your own steer_map.csv file.