Operations Management Assignment

To: 401 Online Students (Do not print your name anywhere)

Subject: Managing a Business - An Excel-Based Dice Game

Purpose: The purpose of this fun and interactive game is to sharpen your understanding of core Operations Management and Theory of Constraints concepts using an Excel-based variant of The Dice Game which Alex played with the boy scouts in Chapter 13 of The Goal.

Story Line: A very close friend of yours, Mr. Herbie, had a great idea and developed a very successful product while pursuing his undergraduate degree in engineering school. Following an advice from engineering school, Herbie decided to open a company, The XYZ Company and made an investment of $50,000 in a production system consisting of 5 processes. His initial thinking was that he will run a single shift operation – 5 days per week, 4 weeks a month i.e., 20 days per month. Each process will have mean daily production capacity of 3.5 units, and the production system should be able to produce 70 units per month (a very definite market demand estimate). Herbie expected mean daily production of 3.5 units to vary + 2.5 units i.e., ranging between 6 and 1 (similar to the fair roll of a dice) due to various sources of variation e.g., machine break down, bad quality raw material, and worker morale. Because of interdependencies among processes and variation, he expected work-in-progress (WIP) to form in front of each process. (Note: Thus, Mean, Variation and WIP levels at each process can be manipulated to represent various states of the production system)

Product Flow: At the 1st process, assume that The XYZ Company has big supply of high quality raw material from our vendor on consignment basis i.e., it belongs to the vendor until the first process uses it. Thus, each process rolls a fair dice on each day which implies that the process has capability to produce that many units (higher number, say 6, represents a productive day and lower number 1 representing a bad day). It also implies that the process tries to produce that many units on that day provided it has work-in-progress (WIP) in its queue from the upstream process. The number of units processed will be the minimum of the WIP available at beginning of that day and roll of a die. The number of units in the queue of a process on any day will be equal to the units shipped from upstream on that day plus ending WIP from previous day i.e., the number of units in the queue minus the number of units produced from previous day. (Note: Units processed at the last process is assumed to be sold, i.e., no finished goods inventory exists).

Performance Measures: How do we know that Herbie understands the goal of the company and he is making right business decisions? Of course, Herbie can look at typical local efficiencies of each process and use bottom line measures e.g., Net Profit (NP) and Return-on-Investment (ROI). Additionally, he might also want to determine the impact of his various decisions on the company’s future profitability by using measures e.g., Customer Service level and Lead Time and company’s operational performance e.g., Productivity and Inventory Turns. Of course, Herbie can also calculate and use TOC measures, Throughput (T), Inventory (I) and Operating Expenses (OE) to evaluate the impact of his decisions on the goal of the company. (Note: Local efficiency of a specific process is calculated by dividing Actual Units Produced by Standard Production Units where the standard is set at Mean Production Rate i.e., 3.5 units per day).

Excel-based Dice Game: In order to use the Excel base-model effectively to run various scenarios, let us explain the model by dividing it into various Panels (You are advised to download and open the Excel Model at this point). Notice that each time you click F9 key on your computer, the results are renewed as a new run. You will use this key, each time you make any changes to the input variables in order to see the updated results.

Panel A is a simple flow chart showing how raw material moves from one process to the next till it is turned into a finished product. The purpose of this flow chart is to be able to create an image of this simple 5 process assembly line and be able to generalize the results to more complex production environments such as Job Shops and large flow shops.

Panel B consists of input variables – Starting WIP, Daily Mean Production Rate, and Maximum Variation around mean - which user can change to represent various scenarios. In the Base Model (see Exhibit I), (i) starting WIP in front of each process is set at zero except for 1st process where we assume unlimited supply of raw material, (ii) Mean daily production capacity is set at 3.5 units, and (iii) Maximum variation around mean is 2.5 units (imitating a roll of dice ranging between 6 and 1). The user is expected to input these variables in this panel to represent various scenarios.

Panel C shows some main results visually e.g., (i) Average Efficiency of each process and Average Daily Ending WIP in front of each process over one month, and (ii) Total Raw Material released in the production system, Ending WIP at each process at the end of 1 month, and Total Units shipped in 1 month. We point out that these results are for one random run representing a single game played for 20 days and a click of F9 (recalculation) key changes these results. More detailed performance measures for the company as a whole are calculated and shown in Panel E.

Panel D shows visually the results similar to Panel C but averaged over 1,000 runs (originally, I had used 200 runs). These results are simulated by using Excel’s “Data Table” function (see Panel F). We notice that with each hit of F9 (recalculation) key, results in Panel C varies significantly where as the results in Panel D do not vary much because these results are averaged over 200 replications. Therefore, these steady state results should be looked and analyzed when we want analyze the impact of a specific scenario where user might change some input variables (see Panel B).

Panel E provides another set of input variables (e.g., assumed monthly demand, days in a month, selling price, raw material cost, period expenses, and investment. Although these input variables can also be varied to create additional scenarios, in this case study we assume these variables are given and fixed. Next, this panel shows the system performance for 1 Run and Average over 1,000 Runs at the end of one month in terms of TOC measures T, I, OE, bottom line measures e.g., Net Profit, ROI, Operational measures i.e., Productivity and Inventory turns, and competitive measures e.g., Lead Time and customer service level. Finally, this panel also shows well known formulas to compute these measures.

Panels F exhibits detailed results of a complete run showing (for each day of the month) the units released in the system, queue length, actual roll of the dice, and units processed at each process respectively. Thus, day by day account of events is recorded for the complete month.

Panel G shows the simulated results for some of the main performance measures and averaged over 1,000 runs using “Data Table” function. Some of these main measures (e.g., Ending WIP and Total Units shipped) are further used to calculate other performance measures over 1,000 runs shown in Panel E.