multi-robot-motion-detection

Instructions:

Create 2 Drive for ros: https://github.com/AutonomyLab/create_autonomy

copy the launch file folder into your ros package.

copy the source,waypoint and destination_robotX to scripts folder of your ros package and make them executable

clone the create 2 drive to your workspace

modify the launch files depending on the location of the scripts.

destination_robotX.py has the destination coordinates for robotX. Modify it to specify it's path

source.py updates the robotX/s_pose from gazebo.

excecute the following on different terminals.

roslaunch [your-package-name] robot_main.launch

rosrun [your-package-name] source.py

rosrun [your-package-name] destination_robotX.py

waypoint.py :

Creates an object of move2goal waits for the rospy to shutdown.

move2goal:

·       s_pose, d_pose topics are created and subscribed to which are of Pose2D() type.

·       an update of d_pose triggers waypoint(s_pose,d_pose)

·       the namespaces for each bot are created in the launch files

source.py :

fetches data from gazebo and publishes it to respective s_pose for all the robots in the simulation
note: this can be modified to launch in a specific namespace

destination_robotX :

contains the destination coordinates as a list
subscribes -> robotX/s_pose
publishes -> robotX/d_pose

drifting is tolerated by sampling the path into smaller lengths of d_sample.

Move away from neighbours:

move_away_from_neighbors/source.py :

similar to source.py described above but it also stores them in a global 2D array (Arr_s) and keeps updating them.

Move away Algorithm:

Arr = Arr_s

r = radius of consideration for individual robots.

Moveaway(robot) {

for robot_i in Arr_s :

    if (distance(robot,robot_i) >= r):

        since robot_i is far; reduce the effect of it on deciding the direction the robot needs to move.

    else:

        since robot_i is too close; increase its effect in deciding the direction the robot needs to move.

Based on the direction, calculate the resultant destination coordinates and return them.

}

move_away_from_neighbors/destination.py:

·       runs move_away_from_neighbors/source.py on a thread to access the updated array

·       parallelly, it implements the “moveaway algorithm” to figure out the destinations for the robots and publishes them to their respective topics until they reach the boundaries.