In this tip we are adding to the simulation from last week’s tip. So we are starting with a simple simulation of a Grocery Store where the activity (All Checkouts) has been set to a Replicate Value of 7. See last week’s Friday Tip for details on how to set this up.
Building the simulation this way means that at a maximum we can have 7 checkouts working at the same time. However the actual number that will be active at any time will be set by the number of available staff as each checkout requires one staff member to operate. So just like in real life even if a store has 7 checkouts, if there are only 3 staff available then only 3 checkouts can be active.
So how can we work out the correct number of staff to use at the Store?
Well, firstly we need to know the parameters; How busy is the store and how long does it take to serve each customer? Lets assume that each work item is one customer; that it takes an average of 6 minutes to serve each customer and that the store is busy with a new customer arriving at the checkouts every 2 minutes.
Setting timing values:
1. Left-click on the start point and in the ‘Properties’, ‘Timing’ section of the ribbon set the distribution type to ‘Exponential’ and the ‘Average’ to be 2 (meaning that a new customer will arrive every 2 minutes).
2. Left-click on the activity and in the ‘Properties’, ‘Timing’ section of the ribbon set the distribution type to ‘Average’ and the ‘Average’ to be 6 (each customer takes 6 minutes to serve).
Testing process capability:
Let’s make one more assumption that the number of staff we will use is going to be 3:
Now run the simulation and keep an eye on the queue just ahead of the checkouts.
During the run you should notice that there is a lot of build up in the queue. To verify this click on the queue, in the ‘Properties’ section open the ‘Results’ option and then check the value for maximum items in this storage. This is the max number of customers who were stuck in the queue at any one moment in time. You will find that this is approximately 70.
Setting a goal and varying the number of staff:
We discover from customer feedback at the Grocery Store that 70 is too high a number to have in the queue. We are asked to hire the minimum number of staff to ensure that the queue does not exceed 15 customers at any time.
We would like to test the simulation to see what happens if we have 3, 4, 5, 6 or 7 staff are available. We could do this by changing the number of available resources and then re-running the simulation for each of these 5 scenarios but fortunately there is also a faster way of doing this; the Scenario Manager.
To use the Scenario manager in this case:
The simulation will quickly run for each scenario and the results of the KPIs will be displayed.
In this case we can see that having 5 staff available will produce a max queue size of 8 customers. So having 5 staff is the minimum amount required to ensure that the queue will not exceed 15 customers (to be sure we could also test this further by using trials).
To see a visual representation of this data: