The autonomous patrolling package

This package allows for the sequential or randomized visit of a set of pre-defined waypoints in the environment, along with a charging behaviour when the battery level drops below a given treshold. It also has two tools for creating and saving a map, and for creating and saving a set of waypoints. It assumes that either a morse simulation or the scitos_bringup have been launched

The map_saver

Without rumblepad

Prior to waypoint autonomous patrolling, a map of the area to be patrolled needs to be created. For that:

1. Have the robot docked (the docking station must be the origin of the odometry frame)

2. Run the map_saver package:

   $ roslaunch map_saver ap_map_saver.launch

The ap_map_saver provides a SaveMap service, that takes as input a string of the form "path to folder where to save the map"/map_name (e.g., /home/strands_user/map) and creates the map_name.yaml and map_name.pgm files in the selected folder.

With rumblepad

The ap_map_saver can also be used in conjunction with the rumblepad, where the user can build a map by driving the robot around and then save it. To do this:

• Have the robot docked (the docking station must be the origin of the odometry frame)

• Before launching ap_map_saver.launch, launch the teleop_app:

     $ roslaunch scitos_teleop teleop_joystick.launch js:=/dev/input/"joystick name"

• Launch the ap_map_saver, with an optional argument:

     $ roslaunch map_saver ap_map_saver.launch map:="path to folder where to save the map"/map_name

• (Optional) Start rviz to visualize the map creation:

     $ rosrun rviz rviz

• Drive the robot around using the rumblepad

• To save the map, press the (A) button.

  • The optional map argument sets the path to save the map. This string has the same format as the one used for the SaveMap service. It's default value is ~/map.

The waypoint recorder

Without rumblepad

Prior to waypoints autonomous patrolling, a file with the waypoints to be visited needs to be created. For that, the waypoint_recorder package can be used:

  $ roslaunch waypoint_recorder waypoint_recorder map:="file path to the map's .yaml file"

The waypoint_recorder provides two services:

• A SaveWaypoint service, that saves the next pose published by amcl in a list.

• A SaveWaypointFile service, that receives a string specifying the file path of the output file (e.g., /home/strands_user/waypoints.csv), and writes all the waypoints saved in the list to that file.

With rumblepad

The waypoint_recorder can also be used in conjunction with the rumblepad, where the user can drive the robot around, create waypoints, and then save them to a file. To do this:

• Before launching waypoint_recorder.launch, launch the teleop_app:

     $ roslaunch scitos_teleop teleop_joystick.launch js:=/dev/input/"joystick name"

• Launch the waypoint_recorder, with an optional argument:

     $ roslaunch waypoint_recorder waypoint_recorder.launch map:="file path to the map's .yaml file" waypoints:="file path to the file where waypoints shoud be saved"

• (Optional) Start rviz to check if the robot is well localized, and give it a pose estimate if needed:

     $ rosrun rviz rviz

• Drive the robot around using the rumblepad

• To save waypoints to the list, press the (A) button. NOTE: The FIRST point you save should be the one in front of your charging station, so that the robot knows where to navigate when it needs to charge.

• To save the list of waypoints to a file, press the (B) button.

  • The optional waypoints argument sets the file path to save the waypoints. This string has the same format as the one used for the SaveWaypointFile service. It's default value is ~/waypoints.csv.

The waypoint patroller

Aunonomously visits a pre-defined list of points randomly or in sequence. Goes to charge when battery drops below a given treshold and after it is recharged, continues the patrolling. Assumes static map and waypoints set are given as input.

• Launch the scitos 2d navigation:

     $ roslaunch scitos_2d_navigation scitos_2d_nav.launch map:="file path to the map's .yaml file"

  • The 2d navigation is kept separated from the autonomous patroller so one can kill the patrolling process without killing the navigation related nodes.

• (Optional) Start rviz to check if the robot is well localized, and give it a pose estimate if needed:

     $ rosrun rviz rviz

• Launch the mongodb_store:

     $ HOSTNAME=bob roslaunch mongodb_store datacentre.launch

  • If you haven't inserted waypoints in your datacentre, you can insert the waypoints from a waypoint log file (as created using the waypoint recorder) into your datacentre:

       $ rosrun waypoint_recorder insert_in_db.py waypoints.csv point_set_name map_name

    Currently map_name is unimportant (until the maps are also stored in db), and the first waypoint is assumed to be the pre-charging waypoint.

• Run the autonomous docking service:

     $ roslaunch scitos_docking charging.launch

  • If this is your first run, or you changed the robot station's location, you need to run the docking calibration:

       $ rosrun actionlib axclient.py /chargingServer

    Then in the Goal textfield complete as follows:

       Command: calibrate
       Timeout: 1000

    Then press the SEND GOAL button.

• Launch the marathon gui, where HOST_IP should be set to an external ip for your machine if you want this to be externally accessible, else can be left blank for 127.0.0.1. E.g.:

         $  HOST_IP=10.0.11.158 roslaunch marathon_touch_gui marathon_gui_dependencies.launch

  • WARNING: If you are running the marathon gui for the first time, before launching, follow the installation instuctions on https://github.com/strands-project/strands_ui

  • After launching, open a browser window and navigate to HOST_IP:8090, where HOST_IP is the one you just set above

• Launch the patroller:

     $ roslaunch waypoint_patroller long_term_patroller.launch waypoints:=waypoint_set_name <randomized:="value"> <n_it:="number of iterations">

  • The optional argument randomized can be true or false. Default is true. If false is given, then the points are visited sequentially
  • The optional argument n_it specifies how many complete iterations of all the points should be done before the patroller outputs succeeded. Default is -1, which means infinite iterations
  • NOTE: For the robot to speak and ask for help, the user that launches this file needs to be logged in the robot's computer. This is related to this
  • If you have the gamepad running, you can pause the patroller and assume control by pressing the dead-man switch. When the dead-man switch is released the patroller resumes patrolling the previous waypoint.

  • You can also pause execution and regain the possibility to control the robot via other sources (e.g. rviz) by running:

       $ rosservice call /pause_resume_patroller

    To resume, just send another service call as above