ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange
Ask Your Question
0

Can i Use navigation stack if my robot does not have any laser or kinect type camera

asked 2018-08-08 11:20:40 -0500

Sufyan gravatar image

currently I am working on a project which is as following. 1-a camera is mounted on cieling of my room 2- image processing is used to localize the robot(using ar alvar track package) and obstacles. 3- path planning is done using A* algorithm i am done with all that is mentioned above. Now i want my robot to follow the planned path. but according to the description of Navigation stack package, you must need a Laser sensor for sensor stream input... Is there anyway to use this package for my project or is there any other solution for tracking of my robot

edit retag flag offensive close merge delete

Comments

Do you have another distance sensor, e.g. sonar?

Humpelstilzchen gravatar image Humpelstilzchen  ( 2018-08-09 13:57:48 -0500 )edit

actually i do have an ultrasonic Sensor mounted in front of my vehicle but it's just to detect if any dynamic obstacle comes up

Sufyan gravatar image Sufyan  ( 2018-08-11 05:04:42 -0500 )edit

You can add a range sensor layer to your local costmap for your ultrasonic.

Humpelstilzchen gravatar image Humpelstilzchen  ( 2018-08-15 10:49:14 -0500 )edit

1 Answer

Sort by ยป oldest newest most voted
2

answered 2018-08-08 19:03:16 -0500

stevejp gravatar image

The navigation stack works perfectly fine without a laser. The main reason you may want to use a laser for navigation (aside from helping with localization) is to mark obstacles on your local / global costmap so that your planners can avoid obstacles. If your global planner ("path planning in A") is already coming up with an obstacle free path, your local planner will just be doing path following with no active obstacle avoidance. However, when you are doing your A path planning, because you won't have obstacle avoidance in your local planner you may want to 'inflate' the area around the detected obstacles (i.e., make the obstacles larger so your global plan isn't super close to obstacles, giving your robot a buffer zone).

edit flag offensive delete link more

Comments

Thanks a Lot!!! i have already inflated the area around the obstacle so as to avoid collision... i was also wondering if I can make a map out of that processed image( containing position of my robot and obstacles) on RVIZ . i am using ar track alvar package to detect my Robot

Sufyan gravatar image Sufyan  ( 2018-08-11 05:04:37 -0500 )edit

Question Tools

1 follower

Stats

Asked: 2018-08-08 11:20:40 -0500

Seen: 386 times

Last updated: Aug 08 '18