0

I am looking for some insight into reporting utilization correctly. I am using a time plot that reports resourceName.utilization(), additionally, I am also adding the utilization values to a Statistics object every hour. I then plot the mean value of this Statistics object as statisticName.mean(). Since utilization in AnyLogic is the returned value is the mean over all individual unit utilization, calculated from the most recent resetStats() call up to current time, does reporting the value statisticName.mean() even make sense? That would be the average of time averaged values.

M Rawat
  • 27
  • 1
  • 7
  • Why not? It gives at high abstraction view. Say, if you have a pool with 3 units and 33% average of averages , it is a reason to investigate it. May be, it has 100 - 0 - 0 load, or 33 - 33 - 33. Both cases require attention and indicate non-optimal workflow. Since AnyLogic allows to get individual statistics, initial 33% is just first high level factor for decision-making. – Gregory Monahov Jan 17 '17 at 13:19
  • Gregory, Thanks for the comment. Maybe I didn't frame my question clearly. Using your example let's say at time 't' all 3 units are 100% utilization. At that point reported utilization will be 100%. For the duration t1 to t2 they are all idle, so at t2 my utilization will be 50%. Furthermore, for the duration t2 to t3 they are again all idle, which would mean my utilization is 33%. Does it then make sense to report the mean of (100, 50, 33) = 61%? I think you meant individual unit utilization which is different from what I am after. – M Rawat Jan 18 '17 at 15:58

0 Answers0