Collect data on anomalies or defects that you have observed. These anomalies or defects are referred to as nonconformities.
A nonconformity can have multiple observations. For example, in January 2016 you recorded a nonconformity against the foundation of a building. The concrete in a specific corner was wet. In July of 2016 you went back to the same place and noticed that the area was still wet, but that it had grown. Both observation records will have different attributes associated. For example, you took a picture in January and one in July as well. Two different pictures attached to two different observations. But both observation records must be recorded against the same nonconformity number. Otherwise, how would you know that they belong together?
To easily differentiate between these two observations, the January record will be Observation 1 and the July record will be Observation 2.