In order for something like this to be possible, we would need to know what the working time is per activity. If it exists as an attribute on the activity table, for example as minutes, the average working time for all activities can be AVG ( "ACTIVITY"."WORKING_TIME" )

In this scenario, the waiting time could be calculated from the throughput time of the entire process minus the time worked. So we would have something like the below:

CALC_THROUGHPUT ( CASE_START TO CASE_END , REMAP_TIMESTAMPS ( "ACTIVITY"."EVENTTIME" , MINUTES ) ) - PU_SUM ( "CASE_TABLE" , "ACTIVITY"."WORKING_TIME")

Hello Tony,

thanks a lot for your answer, I really appreciate it. Unfortunatly I don´t know how to find out if there is a specific defined "worktime". I use the P2P Demo version of Celonis itself and I couldn´t find anything which shows something like the worktime. I just find the start and endtime... in the picture below you can see it with the Variants as dimension.

Hello, I want to explicitly confirm something, because the screenshot is not showing what I expected to see.

Specifically in the Activity table, there is a start time (as is necessary for the existence of the activity table, typically called EVENTTIME but could be called something else). Is there also an End Time?

If so, you can leverage the DATEDIFF() function to calculate a working time, and plug the function into the PQL I suggested above.

To clarify, based on the verbiage, the table you are showing me shows 2 timestamps. The first one is the start time of the first activity associated with the case record. The second one is the start time of the last activity associated with the case record. In order to get a Working Time vs Waiting Time analysis, we must know the start and end time of each individual granular activity, OR the start time and how much time spent working.