OR/MS Today - April 2002 Issues in Education OM and 'Vegas Night' By Arvid C. Johnson In the February column, I discussed an in-class exercise based on the "matchstick game" simulation described in "The Goal" [1]. In this column, I will describe the after-exercise evaluation of the results that I use to "concretize" some key principles of process design and to demonstrate the importance of process flow control mechanisms. Reviewing Students' Expectations Prior to beginning the simulation, students are asked to estimate the number of units that would be completed over a 10-period run for each of the process designs shown in Figure 1. Students' expectations for lines 1 and 5 are typically about 35 units — given that the lines are balanced with average operating capacities of 3.5 units/period — and, similarly, about 70 units for line 4. Students usually identify that the performance of lines 2 and 3 will be limited by the lowest capacity operations at about 35 units. Figure 1: "Design capacities" of the five manufacturing lines used in the simulation. The numbers (1 or 2) associated with each operation refer to the number of dice to be rolled to determine the capacity of that operation. Basic Analysis At the end of the simulation, each line's accountant/scorekeeper summarizes the number of units delivered, WIP in each queue and exposure to defective raw material. We compare the lines' relative performance along these dimensions and vis-a-vis the students' original expectations, which are generally on the optimistic side. Evaluating the available capacity (i.e., "rolled") versus the actual performance (i.e., "moved") of each line's operations will concretize the one-way impact of variability on the lines' operational effectiveness — which can be reinforced through a review of how "slacks" and "gluts" of material moved through the lines. As the discussion progresses from analysis (through "the search for blame" stage) to proposing improvements (i.e., "lessons learned"), I find it effective to structure the discussion around reviewing process design-related concepts versus previewing process management-related concepts. Reviewing Process Design-Related Concepts Process Control: It is important that students recognize that, even though the lines' operations are statistically "in control," their process capabilities may be less than desired. Some users have students use the flip of a coin — corresponding to 3 or 4 units capacity — to demonstrate the impact of improved process capability in a follow-up simulation [2]. Others demonstrate this through the use of altered dice — each set of which averages a given number of units per roll but with different levels of variability [3]. I have students compare the degree to which lines 1 and 4 achieved their expected capacities, since line 4's relative level of variability is substantially less than line 1's. Process Complexity: The relative performance of lines 1 and 5 illustrate the impact of process complexity on performance and, hence, the importance of design simplification. Specifically, I ask students to comment on the degree to which a one-operation line versus a 100-operation line would be expected to achieve their "designed" capacities. Bottleneck Effects: The results of lines 2 and 3 can be used to assess the impact of the bottleneck's location within the process flow. Specifically, I have students compare the performance of line 3's operations before and after the bottleneck operation. Having determined the "best" location for a bottleneck, I ask the students to comment on how they might use this to "manage" the flow of units. Previewing Process Management-Related Material Operating Policies: With respect to operating policies, students often propose guidelines for limiting the WIP in the queues — which, then, leads to a discussion of activating versus utilizing resources [1]. In these discussions, I explicitly address the flow of information within the line versus the physical flow of the units. Flow Control Mechanisms: To preview upcoming material on flow control mechanisms (e.g., material requirements planning (MRP), just-in-time (JIT), theory of constraints (TOC) and "push" versus "pull"), I discuss how each accounts for statistical fluctuations and dependent events and, thus, would alter the simulation. Basic MRP releases (pushes) material into the line on a pre-determined schedule, with the impact of variability occurring as in the original simulation. Thus, achieving better performance would require "closing the loop" (feeding back interim results). JIT systems account for variability by signaling the need for units to satisfy downstream operations as customer demand pulls material through the system. Variability is "reduced" by producing only what is required to meet the demand, often at a rate considerably less than an operation's available capacity. TOC focuses its attention on the bottleneck operation. All variation after that operation is determined solely by the bottleneck's variability (as in line 2), and prior operations are constrained to produce at a rate no greater than needed by the bottleneck. Time permitting, I have even used "Vegas Night" to introduce the use of computer-based simulation in modeling processes, but that's another column for another day. References Goldratt, E., and J. Cox, "The Goal — A Process of Ongoing Improvement," 2nd revised edition, Great Barrington, Massachusetts: North River Press, 1992. Ammar, S., and R. Wright, "Experiential Learning Activities in Operations Management," International Transactions in Operations Research, Vol. 6, No. 2, March 1999, pp. 183-197. Also, see http://webserver.lemoyne.edu/~wright/goldratt.htm Tommelein, I., D. Riley, and G. Howell, "Parade Game: Impact of Work Flow Variability on Succeeding Trade Performance," Proceedings of Sixth Annual Conference of the International Group for Lean Construction (IGLC-6), August 1998, pp. 14. Editor's Note: "Issues in Education" is a regular column sponsored by INFORM-ED, the INFORMS Forum on Education. The column provides educators with practical, useful and thoughtful ideas as they relate to issues in OR/MS education. Educators interested in contributing to the column should contact the column editor, Robert Nydick of Villanova University, at: robert.nydick@villanova.edu Arvid C. Johnson (ajohnson@email.dom.edu) is an associate professor of management in the Graduate School of Business and Information Systems at Dominican University in River Forest, Ill. Table of Contents OR/MS Today Home Page OR/MS Today copyright © 2002 by the Institute for Operations Research and the Management Sciences. All rights reserved. Lionheart Publishing, Inc. 506 Roswell Rd., Suite 220, Marietta, GA 30060 USA Phone: 770-431-0867 | Fax: 770-432-6969 E-mail: lpi@lionhrtpub.com URL: http://www.lionhrtpub.com Web Site © Copyright 2002 by Lionheart Publishing, Inc. All rights reserved.