[uibk_robot_driver] Airskin topic needs to be listened to about squirrel_driver

[shangl]

@qusaisuwan assigned to you ;)

[shangl]

@qusaisuwan behaviour of /airskin/arm_bumper : - sending false all the time; - if airskin is pressed, it is set to true until pressure is gone. up to discussion what happens after it

@bajo @m312z @mzillich @Senka2112 @lokalmatador @federico-b @pregier

[mzillich]

This behaviour (sending false if not pressed) is the same as the bumper ring (topic /bumper).
This behaviour is also what the previous arm control software expected. It makes sense: if you do not get "false", the airskin node might have died, or simply has not been started. So expecting continuous "false" messages is safer.

Hi all,
sorry, I was on holiday and just saw the discussion. I agree with Michael.
The bumper is handled in RobotinoSafety.cpp, I remember I added a piece of code that republish cmd_vel if the bumper is not activated, otherwise it publishes velocity 0. We can do the same for the airskin. Should be 3 lines of code.

[bajo]

Hi. I just wanted to get work on this issue started again. @mzillich and I noticed yesterday that the arm stops as expected when the airskin is pressed. However if it is released again we need to set the mode via rostopic pub /real/robotino/settings/switch_mode std_msgs/Int32 "data: 10" to make it move again.
Otherwise it is stuck, no matter which command we send. Can somebody from Innsbruck have a look into this?