Is it possible to change the input of the sensors in order to test what autonomous vehicles might face in case one or more of their sensors malfunction? Would this effect the way the autonomous vehicle behaves in the simulation or is this something pre-defined?
Hi @xrispa, CRLA does not have a “malfunction” functionality, you can change the parameters of the sensors but nothing else. However, you can implement the functionality you want using the PythonAPI by modifying the information that comes from the sensor before passing it to your driving algorithm.
@Axel Thank you for your reply! I have one more question, besides the sensors, is it possible to alter the status of traffic lights on the runtime and does that affect the simulation?
Sure, you can change the parameters of the traffic lights such as change the time of every color. If you want, you can freeze the traffic light state completely and control them yourself by setting manually the state of the traffic light through the API.
@Axel thank you for your help!!
I have another question about the sensors and their use in the simulation. I found that there is another “module”, ROS bridge that can work with Carla.
Could ROS bridge be used in order to alter the input of the sensors and change the way that the simulation works? Is it possible to give commands to the simulation using ROS bridge that will alter the driving behavior, for example stop the vehicle or turn?
The most up to date docs we have about the ROS bridge is in the repo README, I suggest you read through it and see if it fits what you need, you can indeed communicate with the vehicle and define sensors via ROS but unless you are using ROS in your project, it may not be useful to you