search

osrf / subt

This repostory contains software for the virtual track of the DARPA SubT Challenge. Within this repository you will find Gazebo simulation assets, ROS interfaces, support scripts and plugins, and documentation needed to compete in the SubT Virtual Challenge.
Other
309 stars 98 forks source link

drone rotors never start turing #669

Open zbynekwinkler opened 4 years ago

zbynekwinkler commented 4 years ago

In simulation eaee7507-5006-4529-a2c0-1f283c9481b0 the drone named A600L never took off.

When analyzing the problem, I was able to confirm that some speed commands got all the way to the process /A600L/ros_ign_bridge_twist - see this excerpt from rostopic_stats_logger.log:

topic: "/A600L/cmd_vel"
node_pub: "/A600L_proxy"
node_sub: "/A600L/ros_ign_bridge_twist"
window_start: 
  secs: 86
  nsecs:         0
window_stop: 
  secs: 89
  nsecs:         0
delivered_msgs: 3
dropped_msgs: 0

Further analysis of the file state.tlog revealed that the drone rotors never started to turn. In the file in the "pose" message there are two poses with the name rotor_0 and different ids - 175 and 224. The rotor with id 224 starts to move at around 92s which directly corresponds to our commands to the other drone in the simulation. The rotor with id 175 never turns (its orientation never significantly changes except some noise).

nkoenig commented 4 years ago

At 1:32 sim time, B10W600R spins up, and A600L never starts. I can't see any problems in the logs. I assume you sent the same command to both robots?

zbynekwinkler commented 4 years ago

I assume you sent the same command to both robots?

The rostopic_stats_logger.log notes (quoted above) that the command from our node has been sent to the bridge an 86s of sim time. AFAIU any twist command (no matter the value) starts the rotors. But yes, we send the same commands - the only difference should be 10s wait for the B robot wrt the A robot.