After completing a number of tests with operator and advocate sign-offs and log entry tagging, we THINK that Approval Processor and the automated programs which annotate an event are working properly.  We have now set the FAR threshold used by Approval Processor to 3.8e-7 Hz, i.e. ~1 per month, for each pipeline.  Therefore, please now take GW trigger alerts seriously, since they will represent apparently significant events found by the low-latency analyses.  Operators (along with remote "EM follow-up Advocates") should review and sign off (OKAY or NOT OKAY) each alert, and the Run Coordinators should seriously consider activating the Rapid Response Teams.  (We will need to see, though, whether the low-latency data analysis pipelines really generate event candidates at the intended rate.)

In truth, we are still in a testing mode now; any alerts generated still will not go out to astronomers for follow up without deliberate manual intervention.  However, in this phase we want to test the software and procedures end-to-end, so please try to follow all procedures as closely as possible if you are on shift.  There should be some hardware signal injections in the near future which, hopefully, will be identified by the low-latency pipelines and trigger this whole process (because, for now, Approval Processor is configured to treat hardware injections like real events).  Seeing this in action should help assure us and the Data Analysis Council that the system is ready to go live.