The Unavailable Seconds (UAS) metric indicates during how many seconds of some interval the service can be considered to have been unavailable. ITU-T Recommendation Y.1563 defines UAS based on the concept of Severely Errored Seconds (SES). A period of unavailability starts with 10 consecutive SES and ends with 10 consecutive non-ES. Those first 10 consecutive SES are part of the period of unavailability. See the diagram below for an illustration.
Image source: http://www.itu.int/rec/T-REC-Y.1563-200901-I/en
Netrounds presents the total number of UAS, i.e. the total number of seconds during which the network has been in the unavailable state during the test or monitoring that is currently running.