This could be useful to spot any missing duplicates left in a reading table when there is some overlap in the timestamp of the readings. This can happen we have to manually retrieve previous data to fill in the holes.
For example, try to create a table that shows the number of duplicates per sensor per month.
Probably the easiest route to do this is create a calculated field from the Data Source tab and specify an IF ELSE block to create a new column. Or you can try to aggregate directly a time dimension as we usually do.