The traditional way of dealing with these events was for an on-call traffic manager to review (with the engineering functions) the likely time normal working could be resumed.