Intermediate Modeling 2

5.6

Modify model 5.2 so that the inspectors overlap for the first hour of the first shift and the last hour of the other shift and compare with the original.

A second Day Pattern was created:

1

1

Two experiments can then be run and then making a comparison between the inpection utilization, as well as TIS and WIP. The default warm-up period and replications were kept, at 100 days and 50, respectively.

The additional overlap does not seem to really contribute greatly to the overall efficacy of the system.

Original

Original

Revised Work Schedule

Revised Work Schedule

5.7

Modify 5.2 to tally the number of inspection failures. If the board exceeds 2 failures, reject the board.

A quick step-by-step explanation for accomplishing this in Simio:

  • Create a ModelEntity integer State variable “NumPlace”

  • Create a Model tally statistics element called “Number of Placements”

  • Set a State Assignment on the Rework server at “Before Exiting” and set the following logic to perform a count on the entity:

Exit Counts Logic

Exit Counts Logic

The Routing Logic of the 3 weighted connectors can now be revised to the following:

Bad - (ModelEntity.NumPlace >= 3) * 0.08

Good - (ModelEntity.NumPlace < 3) * 0.66

Rework - (ModelEntity.NumPlace < 3) * 0.26

Ultimately, we find 24 total bad parts created, with 909 good parts.

5.8

Develop a queueing model to estimate fast, medium, and slow-pitch placement machines (38%,33%,29%, respectively).

Count Labels

Count Labels

Using server status labels, we can track the counts of entities (Expression: Processing.NumberEntered). In this case we recieve the following values:

fast<-13333
medium<-11951
slow<-10143
total<-fast+medium+slow
#fast %
fast/total
## [1] 0.3763514
#medium %
medium/total
## [1] 0.3373416
#slow %
slow/total
## [1] 0.2863071

5.9

Pharmacy Model

Setting the model up:

  • 3 Tables
  • RXData (columns RxType, RxMix)
    • Customer = 41%
    • Fax = 59%
  • RXSequence
    • Input at Pharmacist (Server)
    • Input at Sink
  • CustomerSequence
    • Input at Cashier (Server)
    • Input at Pharmacist (Server)
    • Input at Sink
  • A Model Process with Set Row
    • Table Name = RXData
    • Row Number = RxData.RxMix.RandomRow

  • A Rate Table

  • Daily Schedules
    • One for Pharmacists
    • One for Cashiers

Setting up the Source with routing logic and rate tables:

Count Labels

Count Labels

Model Run

Model Run

Results after 500 replications:

  • Customer Time in System : 21.24 minutes (average)
  • Time to fill a faxed script : 13.83 minutes (average)
  • Cashier utilization : 25.6%
  • Pharmacist utilization : 60.19%

Some apparant issues with the model - I had a difficulty properly setting up a balking system, as well as creating a filled prescription resource object (a bin). Adding these would aid in the ability to develop this model.