Enable autonomous driving mode for ego vehicle using C++ about dreyevr HOT 5 CLOSED

I think there might be some confusion here that I'd like to clear up.

The current implementation of DReyeVR "Autopilot" uses the same SetAutopilot function that CARLA uses for their NPC agents. At its core, this uses CARLA's TrafficManager which acts like a simple PID controller for throttle/steering that follows a predefined route. To our knowledge the most straightforward (and intended) way to enable and interact with the TrafficManager was through the PythonAPI as we do in DReyeVR_AI.py. There is no technical limitation as to why you can't implement the interaction to start and communicate with the TM in C++, we just haven't done so yet.

On the other hand, the "handoff" is what we define as the action which switches who/what is in control of the EgoVehicle. By default the user (human) is in control and can drive with keyboard/logi-wheel, but as per our documentation you can also (through C++ events) execute this handoff at any time and pass over control to the Autopilot (assuming the TM is running).

from dreyevr.

Hi! Thank you so much for the quick response. I am quite new to Carla. Could you please direct me to the documentation for setting autopilot using Traffic Manager? I would really apprecaite it.

from dreyevr.

Well the TL;DR of it that we use is to first enable a TM instance in the PythonAPI. Then with that running we can just call SetAutopilot(true) in the EgoVehicle.

For more in-depth documentation here is CARLA's docs.

from dreyevr.

If you are more interested in the innards of the wheeled-vehicle AI controller check out CARLA's WheeledVehicleAIController class for this. (Note that our EgoVehicle's default controller inherits from this class so we have all the same functionality.

from dreyevr.

This helps me a lot. Thank you. I will close this issue soon once I am able to implement it :)

from dreyevr.