The Observation Unit Set (OUS) used by ALMA servers a dual purpose. It provides an organizational structure for separating Scheduling Blocks (SBs) for different configurations or arrays when generating those SBs from the higher level Science Goals defined in the proposal. An OUS also provides an organizational structure for data processing which involves more than one EB, which is the standard case for the ALMA pipeline (unlike the EVLA, where each SB is calibrated separately).
ALMA creates a 3-layer structure of OUSs by default, but only uses the lowest level (also called a Member OUS or MOUS) for any automated processing. The upper layers were envisioned to be used in more complicated imaging or analysis tasks, but they are not currently used, and are not likely to be used in the near future (~10 years). For the time being, we can restrict ourselves to a single OUS layer for any processing we might need.
An OUS is a container at the metadata level, largely analogous to our filegroups for organizing collections of files which make up the contents of NGAS. An OUS links the data used in processing, to the results of that processing. For instance: every science observation by the VLA results in a single Execution Block (EB) which is then calibrated. This can be represented as an OUS with a link to the row in the execution_blocks table and the (yet to be defined or populated) calibration_tables table. A separate run of the calibration pipeline of the same data is represented with a new OUS pointing to the same EB, but to the new calibration information.