ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange
Ask Your Question
1

ROS navigation stack with dynamic map

asked 2017-06-21 11:37:18 -0500

eric1221bday gravatar image

Hi

I'm currently developing an autonomous MAV with the goal of achieving autonomous navigation and obstacle avoidance capability within an indoor environment. With an on board lidar and a downward facing camera using the google cartographer package I have been able to successfully localize the MAV as well as generate an updating 2D occupancy grid. Now the next step local navigation and path planning. From what I've read so far however, the ROS navigation stack, as well the vast majority of path planning and navigation packages out there assume an already generated map for navigation. In my case however I would require that the path planner be able to respond to map changes as well as perform some obstacle avoidance based on the current lidar reading. Are there currently any ROS packages that can accomplish this?

Failing that, are there any literature on path planning on dynamic map coupled with SLAM, so that I may implement it?

Currently the environment is assumed to be hallway like, so a 2D navigation algorithm would suffice.

edit retag flag offensive close merge delete

Comments

I am also working on a similar project. It will be a project that minimizes traffic congestion. However, in this project, a static mapping is always done in navigation and slam applications like your curiosity. Multiple vehicles should recognize each other dynamically and should not collide with each other on the way to the destination. Do you think this is possible? Were you able to realize your project in this way? I would like to review your project.

berk ince gravatar image berk ince  ( 2022-05-29 14:59:42 -0500 )edit

1 Answer

Sort by ยป oldest newest most voted
0

answered 2017-06-22 04:31:15 -0500

saken gravatar image

updated 2017-06-22 04:32:01 -0500

move_base package works well with both planning and obstacle avoidance, the only things needed are proper configuration of the package and giving a proper goals which are reachable. If the goal is unreachable, robot will be stuck trying to reach it.

edit flag offensive delete link more

Question Tools

Stats

Asked: 2017-06-21 11:37:18 -0500

Seen: 1,474 times

Last updated: Jun 22 '17