Skip to main content

Resource utilization looks lower than expected

Symptom​

The resource statistics of a run show low working time or high idle time for resources you know are busier in reality.

Cause​

The simulation measures a resource only against its modeled schedule and the work modeled in this process. Time the real people spend on other processes, meetings, and interruptions isn't in the model, so the simulated utilization reads low. If a resource realistically spends 35% of its time elsewhere, you wouldn't want this process to use more than 65% of it — yet the results page shows that 65% as the resource's entire workload.

Resolution​

Account for the unmodeled work in one of three ways:

  1. Adjust your target instead of the model. Track how utilized the resource is by other processes and treat the remainder as the ceiling for this process — export the results and calculate in a spreadsheet.
  2. Reduce the resource Count. If 7 of a group of 20 are effectively busy elsewhere at any time, model the type with a Count of 13.
  3. Model the interruptions. Give the interrupting work its own generator and activityActivity A named step in a process (e.g. "Approve Invoice"). Each event in an event log records one activity occurrence for one case. that consumes the resource — for example, an Interval generator producing BetweenNorm(3,5) one-hour meetings per week, spread randomly and forced into the active schedule. Stack patterns like this until the planned unavailability is covered.