We implemented a version of the PRMI to DRMI transition that Stefan did by hand 20698 a few weeks ago (22472).  It has worked a few times and not worked a few times, operators are using this only when the DRMI alignment is bad to improve alignment in PRMI. Here are plots of two examples of a sucessfull and unsucesfull transition.  (failure at 20:37:19 UTC today, Oct 27,  sucess at 10/26 23:35:05 UTC )  

In the failure, the PRMI lock was lost when the SRM suspension started to realign SRM, before any feedback came on.  Jeff and I changed the ramp time for SRM in the guardian to 10 seconds (I think the default is 2 seconds).  We will see if this helps the PRMI lock to survive realigning the SRM.  

Also, in the sucessfull attempt the SRCL feedback came on at a time that AS90 was high, and POP90 was low.  It might be that triggering the SRCL output on one of these channels would be better than POP18.