Ask Your Question

Revision history [back]

click to hide/show revision 1
initial version

std::list<navigation goal=""> based navigation

Hi,

ROS-beginner here.

We have an application that sends navigation goal points to an ROS node. This ROS node gives navigation goal points to the robot by a move base action client. What we would like to achieve is to give a list of navigation goal point to the move_base. That is, because of the following: We would like to make our robot move on a specified trajectory. Our 'server' application has methods to break down trajectories into checkpoints. For example: move forward one meter on the x axis - output can be 10 (or more) navigation goal points where x coordinate increments by 10 centimeters. We can give this navigation points to the move base - also calculate estimated orientation - which works well, and the robot moves from checkpoint to checkpoint. Problem is that the robot stops at every checkpoint. We don't want that. We want the robot to follow the goal points one by one without stopping, for example from an std::list, until it's empty. It is acceptable for us to have feedback only at the end of executing a goal list. For now we use the action client, that can send a goal, and wait for feedback. Is it possible to modify the underlying system to accept a list of goals? Do we need to create our own global planner? What would be the easiest solution to our problem?

Thanks

std::list<navigation goal=""> based navigationReaching navigation checkpoints without stopping between them

Hi,

ROS-beginner here.

We have an application that sends navigation goal points to an ROS node. This ROS node gives navigation goal points to the robot by a move base action client. What we would like to achieve is to give a list of navigation goal point to the move_base. That is, because of the following: We would like to make our robot move on a specified trajectory. Our 'server' application has methods to break down trajectories into checkpoints. For example: move forward one meter on the x axis - output can be 10 (or more) navigation goal points where x coordinate increments by 10 centimeters. We can give this navigation points to the move base - also calculate estimated orientation - which works well, and the robot moves from checkpoint to checkpoint. Problem is that the robot stops at every checkpoint. We don't want that. We want the robot to follow the goal points one by one without stopping, for example from an std::list, until it's empty. It is acceptable for us to have feedback only at the end of executing a goal list. For now we use the action client, that can send a goal, and wait for feedback. Is it possible to modify the underlying system to accept a list of goals? Do we need to create our own global planner? What would be the easiest solution to our problem?

Thanks