I am trying something with the simulation of navigation with Turtlebot but with different frame_id's and topics. I checked that the laser data is being relayed to /scan and that my TF tree is consistent and that amcl and move_base are using the right frame_id's and subscribing to the right topics. Even with those conditions and with the laser detecting obstacles, my local costmap is empty (full of zeros) but the global costmap is just fine. Is there some other parameters I should consider that could influence the local costmap?
↧