You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Right now the RLidar is returned with absolute coordinates and always from the same absolute orientation. Knowing this makes it easy to use, but I think it is impractical and unrealistic.
Proposal
It would be a good idea to return it in relative coordinates and taking into account the orientation of the vehicle.
Alternatives
For now I made a prototype inside smarts/env/gymnasium/wrappers/<agent>.py, I would like to add it where the RLidar rays are launched but from there I have not been able to access the agent's orientation.
Right now the logic is simple, first you pass them to relative coordinates and then using np.roll you move the array to match the orientation of the vehicle. This new array is stored in the observations dictionary and now the lidar measurements are 'filtered'.
Additional context
No response
The text was updated successfully, but these errors were encountered:
Motivation
Right now the RLidar is returned with absolute coordinates and always from the same absolute orientation. Knowing this makes it easy to use, but I think it is impractical and unrealistic.
Proposal
It would be a good idea to return it in relative coordinates and taking into account the orientation of the vehicle.
Alternatives
For now I made a prototype inside
smarts/env/gymnasium/wrappers/<agent>.py
, I would like to add it where the RLidar rays are launched but from there I have not been able to access the agent's orientation.Right now the logic is simple, first you pass them to relative coordinates and then using
np.roll
you move the array to match the orientation of the vehicle. This new array is stored in the observations dictionary and now the lidar measurements are 'filtered'.Additional context
No response
The text was updated successfully, but these errors were encountered: