4 Replies Latest reply on May 1, 2017 2:01 PM by Intel Corporation

    looking for robot path planning implementation using ZR300. Suggestions?

    RoBoris

      Hi Developers,

      I am in my plan step to build robotic assistant using Turtlebot as a mobile link and real sense zr300 as camera link.

      The plan is to create a robot which will be able to autonomically map a room and to use it to navigate the room when getting a target point (x,y) on the map.

      I also desire the robot will "know" to avoid obstacles during it's navigation from point A to point B. 

      From reading Intel realSense SLAM Developer Guide I understood that ZR300 provide 2D pose and map as an Output. It also has the ability to make localization pinpoint the robot on the map and make Relocalization to correct the drift in the 6doF.

      To get this outputs i will use the next API's :

      1) realsense_camera - ROS Wiki

      2) realsense_samples_ros/realsense_ros_slam at master · IntelRealSense/realsense_samples_ros · GitHub

       

      Now we got to the path planning step...

      Hasn't found much information about it . I found 2 implementations to perform path planning and recovery behaviors: 

      1) pathplanning/pic at master · pancx/pathplanning · GitHub .

      I see it is a new commit which was not published in the official RealSense GitHub page.

      2) ROS navigation stack which uses move_base .

      1) Do you suggest to use the Ros Navigation stacImplementation? ? Will this module be the right choice to use it with ZR300 when Relocalization (AMCL) and SLAM is done internally inside the RealSense camera?

      2) Does anyone tried using the  pathplanning/pic at master · pancx/pathplanning · GitHub . ?

      3) Do you have other suggestions? implementations? links to a source code?

       

      Thanks !