Hello,

I am trying to develop a solution for autonomous indoor navigation. I am currently simulating it on Gazebo and using these elements:

• hector_quadrotor for the simulation in Gazebo
• hector_mapping for localization
• move_base for path planning

My problem is the link between my robot model (the baselink frame) and my scanmatcherframe, created by hectormapping and supposed to be my odom frame. The link between the two parts should be done by the hectorpose_estimation but I am facing 2 problems:

1) Should the nav frame of poseestimation be transformed into the scanmatcherframe directly?

2) The height sensing of the poseestimation doesn't provide good results, how can I use the sonarheight information instead of the altimeter?

Is there simple ekf nodes to combine 2D pose estimation with height sensing in order to return a simple 3D pose?

Thank you for your help,