Welcome to Communities.
I see you have '13 of 13' disks in use but '12 of 12' on the host below it - does the host with the device that went PDL have 12 disks + one PDL reference being counted or actually 13 devices?
(e.g. the disk may have come back online but now how has 2 references with different UUIDs but only one with naa)
If it is actually kaput then you likely won't be able to remove the PDL reference with anything other than 'No data evacuation' option - only do this if health check is green for
Cluster > Monitor > vSAN > Health > Data (and you have no missing FTT=0 data if using such policies).
If it doesn't complete with this option (as it indicates), you can try removing the reference via CLI:
# esxcli vsan storage remove -u <UUID>
NOTE: the above is a destructive command, if used improperly can cause data loss (as noAction is default) - ensure you refer to the correct device and that data is otherwise healthy.
Either way, you should have a look in vmkernel and vobd.log for more information on cause of the device failure unless it is already apparent from other means (iLO/iDrac, disk LEDs, no power).