Set up performance measure tables

00:03

Performance measure tables enable you to systematically track, measure, and analyze specific output values for your FlexSim Model.

00:12

Examples of output values are total throughput, average wait time, or maximum key length.

00:19

In this example, FlexSim is open to an already created model.

00:24

To add a Performance Measure Table, in the Toolbox, click Add, and then select Statistics > Performance Measure Table.

00:34

In this case, dock PerformanceMeasures1 next to the Model so that you can view them side-by-side.

00:42

From the Toolbox, you can add multiple tables by right-clicking Performance Measure Tables

00:47

and selecting Add Performance Measure Table or by following the previous steps.

00:53

For this example, use the table you just added.

00:57

Here, in PerformanceMeasures1, each row is a performance measure, and you can use the up or down arrows to adjust the number of rows.

01:07

Keep in mind that each Name must be unique across all performance measure tables for your model.

01:13

For this example, rename PerformanceMeasure1 to “Throughput”.

01:19

Click in the Value column, and then expand the drop-down to access the Value Properties.

01:25

Here, select the Reference Sampler.

01:29

Then, in the Model, click Processor1, and in the menu that opens, select Statistics > Output.

01:40

Back in the table, use the Display Units and Description fields to organize and reference your performance measures.

01:47

For example, enter “items” for the Display Units and “Number of items completed” for the Description.

01:56

Reset and Run the Model.

01:59

When an item moves through the process, the Value in the Throughput field increases.

02:04

Here, you can see that the value first increases to 1, and then to 2 as the simulation is Stepped forward.

02:12

Stop and Reset the simulation.

02:15

The Reference Sampler allows you to sample not only most 3D objects in your Model,

02:20

but also most activities and shared assets in your process flow, groups in your Toolbox, and charts on your Dashboard.

02:28

For example, to review the travel distance of the operator, expand the Value field for PerformanceMeasure2,

02:36

click the Reference Sampler, and then sample Operator1 in the Model.

02:42

In the menu that opens, select Statistics > Travel Distance.

02:49

In the Table, name the performance measurement “DistanceTraveled”,

02:54

update the Display Units to “meters” and add a Description of “Distance traveled by operator”.

03:01

Next, sample a Group.

03:05

In the Model shown, the three processors are already part of a group called Processors.

03:11

Expand the Value Properties for PerformanceMeasure3 to select the Reference Sampler.

03:18

Then, in the Toolbox, under Groups, select Processors.

03:25

In the Value Properties, expand the Value field and select State percentage by group.

03:32

Then, set the State to 2 – processing and the Aggregation to Average.

03:40

Back in the Table, name the measure “ProcessTime”,

03:45

set the Display Units to “seconds” and add a Description of “Average time spent processing among group”.

03:52

When you Reset and Run the model, you can see the values update.

03:57

Note that, in this example, the ProcessTime measures the processing time in the model shown here,

04:03

while the DistanceTraveled measures the operator in the model above.

04:07

These are just a few examples of performance measurement within your model;

04:12

there is tremendous variability in how you can set up performance measure tables to track various aspects of your model.

04:19

Keep in mind that performance measurement can require substantial computational capacity, as these values are constantly being updated.

04:27

If you find that your model runtime is slow,

04:31

best practice is to close the Performance Measures pane and reopen it once your model finishes running.

04:37

You can also use performance measures in conjunction with parameters and the experimenter tool

04:42

to systematically track, analyze, and compare key performance metrics across different simulation scenarios.