Academic Integrity: tutoring, explanations, and feedback — we don’t complete graded work or submit on a student’s behalf.

Kits of ceiling fans arrive at an assembly system with TRIA(2, 5, 10) interarriv

ID: 3686533 • Letter: K

Question

Kits of ceiling fans arrive at an assembly system with TRIA(2, 5, 10) interarrival times (all times are in minutes). There are 4 assembly operators, and the kits are automatically sent to the first available operator for assembly. The fan assembly time is operator-dependent as given:

Operator 1             assembly time: TRIA(15, 18, 21)

Operator 2            assembly time: TRIA(16, 19, 22)

Operator 3            assembly time: TRIA(16, 20, 24)

Operator 4            assembly time: TRIA(17, 20, 23)

Upon completion of the assembly process, the fans are inspected with approximately 7% being found defective. A defective fan is sent back for repair to the same operator who assembled it. These defective fans have non-preemptive priority over incoming kits. Since the fan needs to be disassembled and then reassembled, the repair time is assumed to be 30% greater than the normal assembly time. Run your model for 20,000 minutes and collect statistics on operator utilization and the time in the system.

Explanation / Answer

a) Suppose you could hire one more person and that person could be a second (identical)

operator at any of the four locations. Which one should it be? Use FAN with five

replications per scenario and select the best in terms of lowest average time in system.

In retrospect, is your choice surprising? How important is it to get this choice right?

b) Now suppose you could hire up to five more people and allocate them any way you

want to the four existing operators, including allocating all five of the new people to a

single operator. What's the best thing to do? Use OptQuest to search for the optimal

allocation of this maximum of five more people, with the objective of minimizing the

average time in system . If OptQuest displays the following error message “The

Arena model has no controls defined. The model must

have at least one control to run optimization”, although you have resources in your

model, you will need to change the Decimal Symbol parameter defined for the

computer to a period instead of a comma. To do this go to Start/Settings/Control Panel

and click on the Regional and Language Options icon and under Regional Options click

on the Customize button, there you will be able to define the Decimal Symbol to a

period "." instead of a comma ",".

c) The quality control staff for the fan assembly area has decided that if a fan is rejected a

second time it should be rejected from the system and sent to a different area for

rework. Make the necessary changes to the original model (the one you built before

exercise a), run the simulation for 20 000 minutes and compare the utilization by

operator and time in system (based on 5 replications of each model). Use the Output

Analyzer in order to compare the alternatives . Is there any statistical difference

between the performance indicators (utilization per operator

and time in system) from the two models? Save this version of your model as

teamXX_task3_1c.doe (where XX is your team number).

Keep track and report the number of fans rejected from the system. (Hint: Set an

attribute for storing the number of times the parts were rejected. Use a “Decide” module

from the Basic Process Panel in order to define the path of the entities basing the decision on condition.)

In the report:

Describe the analysis process in each of the subtasks and write down the results of each experiment

you do and discuss them. Include screenshots from Output Analyzer, Process Analyzer and

OptQuest.

Models:

teamXX_task3_1.doe

teamXX_task3_1c.doe

(where XX is your team number)