search

rosin-project / metacontrol_sim

2 stars 4 forks source link

Data logging tool for the experiments #25

Closed chcorbato closed 4 years ago

chcorbato commented 4 years ago

This is needed for #26

marioney commented 4 years ago

@chcorbato @darkobozhinoski what data should we log?

I'm listing here some, but please feel free to add / remove anything that may be needed

darkobozhinoski commented 4 years ago

@marioney , we will also need to log the time. With this, we can measure the total time to complete the mission, the time that certain attribute was under a safety/energy threshold etc. @marioney , we should also measure how many times reconfiguration have been mane (we purposely killed the move_base). The rest looks good to me.

chcorbato commented 4 years ago

I'm listing here some, but please feel free to add / remove anything that may be needed

Maybe it is handy to log when the nav goal is sent and when it is achieved, I think /move_base/goal and /move_base/result\ I already added this to the list

@marioney , we will also need to log the time.

ROS messages (whet we will log), include a header with a timestamp See #30 for details

gavanderhoorn commented 4 years ago

ROS messages (whet we will log), include a header with a timestamp

Nuance: only those messages which have a std_msgs/Header.

You don't get anything for free here.

If you don't have a header, you'll only be able to approximate things with ros::Time::now(), which is suboptimal.

chcorbato commented 4 years ago

I think for the moment this is implemented ad-hoc for our current experiments in #26 I'm triaging it as enhancement and for later in our project board