OM Simulation Descriptions and Implementation Tips. After completing the simulation, capture a screen image of your final simulation results including the rubric evaluation metrics (i.e

Operations Management Forecasting

Holly Crosley

Colorado State University Global

OPS510: Operations Management

Dr. Parimal Kopardekar

November 28, 2021

1

2

Operations Management Forecasting

Forecasting is taking historical data to make future predictive assumptions. Managers can

use forecasting to estimate orders for future goods and services. Accurate forecasts help

businesses reduce expenditures for raw materials. Proper scheduling of staff is an additional

benefit. (Heizer, et al., 2020) It can be considered an artistic science (Trevidi, 2017).

Makridakis, et al. (2020) make it clear when forecasting there is no way for anyone to know with

absolute certainty a predictive outcome. Over the years many different forecasting methods and

techniques have been developed and tested. Regardless of which method or technique is

employed, the analyst should remember past results are not necessarily an indicator of future

events (Seroney, et al., 2019).

Many events can affect the efficiency, quality, profitability, and customer experience

adversely. Supply chain interruptions can be due to supply, demand, transport, volatility of the

market, and political unrest. The petroleum industry is a large and complex industry where quick

and accurate forecasting is incredibly important lends itself to demand forecasting. (Seroney, et

al., 2019)

PetroPlex is a fictional company in a simulation from Pearson as part of the Operations

Management book by Heizer, Render, and Munson (2020). The simulation scenario results will

be discussed as they relate to forecasting. There were different forecasting methods and

techniques utilized to make the forecast. Following the simulation, upon reflection, lessons

learned were applied to the author’s analysis.

Forecasting Simulation

The forecasting simulation was 24 month simulation testing demand predictions for

gasoline. The premise of the simulation was:

3

You have taken up the role of an Operations Consultant and just signed a 2 year contract

with a new client, PetroPlex, which is a gas station in your area. As part of your role, you

will provide monthly forecasts for PetroPlex to match customer demand. PetroPlex sells

3 types of gas – Regular, Midgrade, and Premium. You will need to provide accurate

forecasts of the demand for the three types of gasoline at the beginning of each month.

Your performance will be based on the collective Mean Average Percentage Error

(MAPE) of the three types of gas. The final MAPE value should be less than 25%.

The simulation began with emails, texts, and voice mail messages from the owner and

subject matter experts setting up the scenario and giving background data about historical

gasoline sales. An expert in gasoline sales forecasting offers his service 3 times over the course

of 24 months. The simulation requires a decision to be made about volume of gasoline for the

next month of sales. Every month following a new email or text would come in and every few

months, the owner would leave a voice mail with either praise or reprimands.

In performing the simulation forecasts, I utilized a hybrid approach. The demand forecast

with naïve forecasting utilizing trend projections and sales force composite. In order to make my

monthly forecasts, I looked at the monthly communication from the SMEs and analyzed the data

they suggested. The messages varied from local information, national information, and global

information. All three had different ramifications in regards to the forecast for that month.

Sometimes the information complemented the trend data and sometimes it was in direct contrast.

The demand trend data was the second item I looked at when considering my monthly forecast.

The table showed the monthly demand from the previous 12 months. This historical data was

helpful to see the purchasing trends of consumers. The other item I looked at and considered was

the small trend picture that was located on the main page of the simulation. Finally, I had to

4

consider whether to use the expert or not. All of this data was reviewed to make a decision for

the monthly forecast (see Appendix B).

Operational Management Forecasting Methods

A demand forecast is one of three main forecasts, along with economic and technological.

Demand forecasts focus on using existing data to quickly respond to what the customer wants

and needs. (Heizer, et al., 2020) The PetroPlex simulation offered lots of historical data and

relevant current event data. And, since the customer request was for a monthly decision in order

to make a quick demand forecast.

A common forecasting model is the time series model. A time series model focuses on the

historic data and trends from that data. (Heizer, et al., 2020) The time series model is the best

model for this PetroPlex simulation based on the plethora of trend data.

One of the qualitative methods used in this simulation was similar to a sales force

composite. A sales force composite is when sales personnel make their projections for a certain

time frame and those projections are combined for one larger forecast. (Heizer, et al., 2020) In

the simulation, it can be presumed all the PetroPlex gas stations send their demand data to

corporate so the complete demand trend can be computed for the total historical trend.

Simulation Results

I completed three total simulations. The first simulation did not go well. My ending

MAPE was around 30%. As I looked at the summary, this was due to a typo around month 12. I

entered 95000 when the forecast was supposed to be 9500. While this was an unfortunate

typographical error, it was a lesson where it could have actual negative consequences if it had

happened in the real world. Thankfully, this was a simulation and the worst thing that happened

is I lost the contract (which would also have happened in real life). It was a great reminder of the

5

importance of double-checking data. In this two-year simulation, the remainder of the contract

(about 12 more months) was not enough to overcome the error.

The second simulation performed went well (MAPE was just barely over 10%), but no

bonus was obtained. I remembered to use the expert in this simulation, but the recommendation

was not what I thought it should be based on the email received from the beginning of the month.

I did not use the suggestion and was more correct in my forecast that the expert would have been

in his.

The third time was best, and the bonus was achieved. I did not use the expert in this

simulation since my forecasts were looking accurate across all three gasoline grades (see

Appendix A). Using the historical data predictions and the subject matter experts (SMEs)

enabled me to determine an accurate forecast for PetroPlex. I achieved the bonus by obtaining a

MAPE of 7.58% (see Appendix C).

Lessons Learned

Lessons learned are captured when the positive and the negative experiences are

documented (Project Management Institute, 2017). In the PetroPlex simulation, there were many

lessons learned. There were 3 main lessons that allowed me to be successful in the final

simulation.

The first was making sure to pay attention to the monthly communications from the

SMEs. They directed attention to what was happening globally with regards to the petroleum

industry. While the SMEs provided some general information, it was up to me to determine what

to do with that information.

Second lesson learned was to continually refer back to the past year’s trends. The

monthly demand trend chart was sent in an email communication at the beginning of the

6

simulation. This was helpful to see what months the demand was higher and when it would

wane.

Remembering and considering the Expert was the third lesson learned. In my second

simulation, I remembered to ask the expert, but his recommendation was solely based on the

demand trend from the previous year. He did not take the SME email message into his

prediction. I chose not to use his recommendation for that month’s forecast. His recommendation

also made me question whether any of his other forecasts would be accurate and correct, so I

chose not to use him. On my third simulation I was able to perform the entire simulation without

help from the expert. If I had questions or was second guessing my forecasts, I would have asked

him.

Conclusion

“Good forecasts are an essential part of efficient service and manufacturing operations.”

(Heizer, et al., 2020) Forecasting is an important step in Operations Management to make

accurate and effective financial decisions for businesses. The science of forecasting is the

techniques used, while the application of the data is the art of forecasting (Trivedi, 2017). There

are many different methods and techniques to use to forecast data. Managers need to make a

decision on which to apply to come to a conclusion relatively quickly. (Heizer, et al., 2020) The

PetroPlex simulation showed how the naïve forecasting method can work in a real-world

situation, by taking news and historical data to make a current forecast about gasoline usage.

Documenting lessons learned after forecasting is a good best demonstrated practice to assist the

manager and additional stakeholders for future forecasting.

7

References

Heizer, J., Render, B., & Munson, C. (2020). Operations management: Sustainability and supply

chain management (13th ed.). Pearson Education, Inc.

Makridakis, S., Hyndman, R. J., & Petropoulos, F. (2020). Forecasting in social settings: The

state of the art. International Journal of Forecasting, 36(1), 15-28.

PETROPLEX, Heizer, J., Render, B., & Munson, C. (2020). Operations management:

Sustainability and supply chain management (13th ed.). Pearson Education, Inc.

Project Management Institute. (2017). A guide to the Project Management Body of Knowledge

(PMBOK guide) (6th ed.). Project Management Institute.

Seroney, J.K., Wanyoike, D.M., Langat, E.K. (2019). Influence of Demand Forecasting on

Supply Chain Performance of Petroleum Marketing Companies in Nakuru County,

Kenya. THE INTERNATIONAL JOURNAL OF BUSINESS MANAGEMENT AND

TECHNOLOGY.

Trivedi, B. (2017, May 3). Demand Forecasting: The Art and Science That Keeps You

Guessing. Arkieva. https://blog.arkieva.com/demand-forecasting-art-science/

8

Appendix A

Screen capture from the Pearson OMLab PetroPlex Simulation which shows the

forecasted versus the actual demand for each of the gasoline grades (regular, midgrade, and

premium).

9

Appendix B

The decision process for making the demand forecast for the PetroPlex Simulation

The screenshots were taken directly from the PetroPlex simulation.

10

Appendix C

Overall MAPE (%) results from PetroPlex Simulation showing previous and current

MAPE (%) from screenshot of Summary on PetroPlex simulation on Pearson MyOMLab.