Force update when there is no movement

[olmerg]

in the parity amcl this is implemented though a service request_nomotion_update. Although in nav2_amcl is just a flag, in beluga it is required to modify the policies like resample_on_motion_policy

Definition of done

• Define the way to be implemented
• implement the service in beluga_amcl
• modify the policies in beluga

[olmerg]

• Request no motion update put in true force_update_. this variable modifies if we should update the filter. Additionally, when this flag is true the particle cloud is not published. Some additional point about this flag is that set to true on_cleanup and set false after it is used to decide read the laser. So conclusion this service forces the filter to work when there is no movement.

[nahueespinosa]

Looks like nav2_amcl doesn't perform sensor and motion model updates if there is no motion.
For reference:

• nav2_amcl/src/amcl_node.cpp#L668-L670
• nav2_amcl/src/amcl_node.cpp#L677-L678

Note that in our implementation we only skip the resampling step if there is no motion, which is not the same.
I don't know how much sense it makes to implement something like this if the underlying behavior is different.

[stwirth]

This feature is useful to refine a manually set pose (via /initialpose), see ros-planning/navigation#1226.

[nahueespinosa]

Thanks for the context! When I commented a few months ago, Beluga was updating the filter motion and sensor models all the time, so a manual update was not required for the robot to re-localize when the initial pose was published.

In the current version, after #233 which makes resampling policies much more similar to AMCL, this makes a lot of sense.

[hidmic]

Solved!