Gerardo Puga

I don't think this is a situation that deteriorated over time from start of year, really, but now we have two independent targets in beluga_amcl that get built at the...

A current thread with a lot of information about this: https://discourse.ros.org/t/openvdb-compile-size/38790/4 > Each build agent is a compute-optimized “xlarge” instance in EC2 spread between c5.xlarge, c6i.xlarge, or m5.xlarge which all...

> A stopgap solution might be to temporarily daisy chain those two targets with cmake so they don't get built at the same time, while we look for a better...

> Hey! I don’t see any data newer than 2023/09/02 in the [reports](https://github.com/Ekumen-OS/beluga/blob/main/beluga_benchmark/docs/reports). Do we have something more up to date? If so, where can I find it? That's very...

I'm still digging into this issue.

The issue is in Beluga, not in the [Lifecle manager](https://github.com/ros-navigation/navigation2/tree/main/nav2_lifecycle_manager) nor the [bond_core](https://github.com/ros/bond_core) library. The reason the lifecycle manager node kills beluga regardless of the `heartbeat_timeout` value that is configured...

This potential problem was atenuated with the fix to #468 , which improved the performance of the algorithm calculating the likelihood map from the occupancy grid of the environment. The...

This was addressed in https://github.com/Ekumen-OS/beluga/pull/473

This is the behavior for both Beluga and Nav2 in the same dataset, but using the Likelihood model: [likelihood_both.webm](https://github.com/user-attachments/assets/9ced4427-d7f4-40a8-8fe6-a98538754b8c)

This is Beluga using the beam model: [beam_beluga.webm](https://github.com/user-attachments/assets/a4e254e9-3801-4ca0-86b0-cc75f881dc6c)