Sequential EntityProcessors, Maintenance, and unexpected Idle Time

[Anthony Earl]

Hi,

  The real-world I'm trying to model at scale are teams planning work and working through it.  They are interrupted by corporate tasks and meetings which I'm modeling as Maintenance on the ResourceUnits (people).

What I'm seeing is that, in addition to Maintenance Time, some people are experiencing significant Idle time which I cannot fathom since there always seems to be plenty of work to do.  

I first saw this in my large-scale model as soon as I added maintenance time.  That was in version 2024-03.  I see exactly the same in 2024-04.

The enclosed simple example is attached.  There's just 15 minutes per piece of work for planning work.  The work itself takes 10 hours.  There's some variance in the frequency and duration of interrupts (maintenance) based on Triangular distributions.

When the Capacity of the PlanWork stage matches the Capacity of the Work stage I see no Idle time for the team members.  Obviously, if the capacity of the Work stage is less than the size of the team, I would expect, and do see, significant Idle time (since not everyone is allowed to work at once).

BUT, when the Capacity of the PlanWork stage (set to less than full team size) is significantly less than the Work stage (with capacity set to full team size), there is always either planning-work or work to do.  So I would expect *no* significant Idle time (perhaps just an hour or so as the first plans complete).  But I seem to see at least one team member with around 20 hours of Idle time and I cannot understand that.

I realize that I can work around this by setting full-capacity at the planning stage but I'd very much prefer to understand what's happening so that I don't accidentally reach the wrong conclusions about the impact of corporate interruptions.

Thanks so much for this great tool and community.

[Anthony Earl]

I've learned a little more about the behavior of the model I referenced.

If I don't use the ResourcePool and simply reference all of the ResourceUnits then I do not see the unexpected Idle Time, even if the Capacity of the PlanWork step is substantially less than the Work step.

This makes me wonder whether there's something about a ResourcePool that is failing to let EntityProcessors "know" that a ResourceUnit is back from Maintenance.