scenario_runner
scenario_runner copied to clipboard
carla-simulator
Using the scenario runner with Carla and autoware
For creating an assessment environment for autonomous driving software, I want to run scenarios with my autonomous driving stack as the agent. Therefore I installed the scenario runner and everything that comes with it and it also works for the "normal" use case of manual control.
But when I want to connect the scenario to autoware, by launching the carla_autoware bridge next to it, the scenario loads the initial configuration and it is also visualized in a pygame window, but there is no reaction in autoware and I cannot launch the automated driving, except with the "autopilot" button P, which I had to learn is not my autonomous driving stack.
I am using this carla_autoware_bridge: https://github.com/Robotics010/carla_autoware_bridge
There is also this documentation on how to setup everything, but where it says to press Button "P" for enabling the autonomous driving software. (https://medium.com/@wuxiaohua1011/wip-carla-autoware-integration-and-scenario-runner-installation-guide-f4c6a803f49a)
Desktop:
- OS: Ubuntu 20.04
- CARLA Version CARLA 0.9.12
- Python version 3.7
- ROS galactic
Hello, which autonomous driving stack are you using? Is it a custom one?
You are correct, you do not want carla autopilot (by pressing P) if you want to use your own stack.
To be able to use autoware I believe you need to publish on the correct ROS topics since autoware is interacting with ROS using these topics. There are some ROS services which will handle publishing data to some relevant topics. There may be some cases where you want to publish to these topics directly from your program. This page may help you understand the autoware API.
No I am just using autoware. So what you are saying is that I have to write my own code for publishing to these crucial topics and there is nothing similar already existing?
Hey @endris1011! I'm currently looking on using scenario runner for autoware and the bridge. Could you describe more in detail what scenario you are running and what results are expected to have?
