EIND 422

Exam 2: Take-home Simulation Exam

Due: June 29^th at 10:00 PM

You need to submit in a copy of the SIMIO file(s) and report in the Exam 2 Assignment Folder.

Exam Policies:

• This exam is to be completed individually. Do not accept or give assistance.

• Use the Simio report guidelines (report template provided in pages 4-6 with grading scheme).
• Submit the electronic copy of the report along with the Simio files (Part A and Part B) in D2L.
• The professor will not be available to answer questions (assumptions, modeling, or results interpretation type).
• Please, express all output time metrics in minutes.

Consider a process with two different types of entities. Assume arrivals for each entity type vary through the day. Table 1 contains the non-stationary Poisson Process arrival rates for entities of type 1 and type 2. Please note that λ₁ and λ₂ are already expressed as hourly rates.Assume all entities travel at the default speed of 1.4 meters/second.

Table 1: Hourly arrival rates (entities/hr.)

Upon arrival, entities type 1 travel 4 meters to Station 1A where they are processed following a triangular distribution with parameters (2, 5, 7) minutes. Similarly, type 2 entities travel 9 meters and arrive at Station 1B where all entities are processed following a triangular distribution with parameters (3, 6, 9) minutes.

All entities (type 1 and type 2) then travel 5 meters to Station 2 where they are all processed following a triangular distribution with parameters (2, 3, 4) minutes.

Immediately after exiting Station 2, entities type 1 return to Station 1A via a path that is 7 meters long whereas entities type 2 proceed to Station 3 through a path that is 5 meters long.

Entities type 1 entering Station 1A for the second time are processed following an exponential distribution with mean processing time of 3 minutes and then exit the system through a path that is 10 meters long.

Entities type 2 going through Station 3 are processed following an exponential distribution with mean 5 minutes. Station 3 involves a machine which occasionally fails. Mean time to failure (up time) follow an exponential distribution with mean of 3 hours; mean time to repair (down time) also follow an exponential distribution with mean of 20 minutes.Once entities type 2 finish processing at Station 3, they exit the system through a path that is 5 meters long.

Station 1A, Station 1B and Station 2 are human processes with varying staffing levels through the day. Table 2 contains the staffing data of these stations through the day.

Table 2: Staffing data

Assume the process starts at 8:00 a.m. and ends at 10:00 p.m. and you can ignore entities that are still in the system at closing time (probably not the best customer service!).

Part A

• Develop a Simio model of this system.Performance metrics of interest include:
• Run the simulation model for 30 replications with 1 hour warm-up period and a 95% confidence level.
• Finally, animate the simulation to make it look attractive.

• average time type 1 entities spend in the system
• average time type 2 entities spend in the system
• scheduled utilizations of the servers at Station 1A, Station 1B and Station 2

Part B

The process engineer is considering firing one of the employees at Station 1B that currently works from 3:00 p.m. to 10:00 p.m. changing the staffing of Station 1B according to Table 3.

Table 3: Proposed staffing data

• Make the changes to the simulation model to study the potential effects of such staffing change.
• Run both models (Part A and Part B) for 250 replications. As before, use 1-hour warm-up period and a 95% confidence level.
• What can you conclude from this change? Will firing the employee affect the system in terms of average time in system for entities type 1 and type 2? What recommendation would you make in terms of firing the employee? What other recommendations would you make?

Final notes:

• Methods section is the same for both parts since the change is just on the staffing.
• Results section should contain:
  • Table summarizing results for Part A (30 replications). Don’t forget the 95% confidence intervals.
  • Table summarizing results for Part B (both systems for 250 replications).
• Address questions above in part A-1 (what did you get for the metrics of interest?) and B-3 in the Discussion section.
• Please, do not forget units and Confidence Intervals! Please, express all output time metrics in minutes.

(0.5) Title

(0.5) Name:

I. Introduction (27)

• (6) Problem description
• (5) Assumptions
• (4) System initial state- number of entities in the system, # of servers, server status
• (4) Table summarizing the current system inputs.For example:
• (8) A flowchart or block diagram of the process. For example:

Table 1- Summary of current system inputs

Figure 1- Generic Process flow chart

Or

Figure 2- Generic Block Diagram

II. Methods (14)

• (10) Detailed description on how you simulated the problem.
• (4) Include 2 print screens for the SIMIO model from part A (part B should look alike); one in 2D and one in 3D. For example:

Figure 3- Model 5-2 in 2D

Figure 4- Model 5-2 in 3D

III. Results (30)

• (10)Table summarizing results for initial system when running both for 30 replications.
• (20)Table summarizing results for both models when running both for 250 replications.

IV. Discussion (20)

• (8) Answer the questions asked.
• (6) Discuss any limitation(s) of the model.
• (6) Also, include any comments or observations you might have.

V. Conclusions (8)

• (5) What can you say from the simulation results?
• (3)If you were a consultant, what recommendations would you make?

Please note that tables are labeled on top and figures are labeled on bottom. Use only the categories of “table” and/or “figure”.Pictures, diagrams, charts, plots, etc. all fall under the “figures” category.

https://textbook.simio.com/SASMAA/index.html