Speech to wav timeout would timeout if someone is speaking for too long

[tomkimsour]

The idea behind having a timeout for the speech recognition is to be able to restart the procedure if the person talked but the robot didn't not record it.
When someone speaks he/she expect a feedback weather the robot detected someone speaking, understood what the person said or did not understand what the person state.
As we have no way to tell when someone actually start speaking, we can't ignore the timeout used to give feedback in case nothing was heard.