redesigning-a-lean-simulation-game-for-more-flexibility-and-higher-efficiency

AC 2012-4003: REDESIGNING A LEAN SIMULATION GAME FOR MORE FLEXIBILITY AND HIGHER EFFICIENCY Dr Hung-da Wan, University of Texas, San Antonio Hung-da Wan is an Assistant Professor of the Mechanical Engineering Department and the Director of the Sustainable Manufacturing Systems Lab at the University of Texas, San Antonio (UTSA) He teaches Six Sigma and lean methodologies, computer integrated manufacturing systems, and manufacturing systems engineering He is among the core faculty of the Center for Advanced Manufacturing and Lean Systems at UTSA and has been offering short courses to the industry His research interests include sustainability of manufacturing systems and web-based applications in manufacturing Miss Yi-Ching Liao, University of Texas, San Antonio Yi-Ching Liao is a master’s student in advanced manufacturing and enterprise engineering at the University of Texas, San Antonio (UTSA) She is also a Graduate Research Assistant at the Sustainable Manufacturing System Laboratory at UTSA She received her B.S in system and naval mechatronic wngineering from National Cheng Kung University (NCKU) in Taiwan Her research interests include lean systems design and implementation, simulation and gaming, and engineering education Dr Glenn Kuriger, University of Texas, San Antonio Glenn Kuriger is a Research Assistant Professor at the Center for Advanced Manufacturing and Lean Systems (CAMLS) and the Department of Mechanical Engineering at the University of Texas, San Antonio (UTSA) He has been involved with lean implementation projects with several organizations He has done lean work in the manufacturing, service, and other industries He is also working to develop and teach a series of lean short courses that CAMLS is currently offering to industry His current research interests include lean concepts, lean healthcare, lean simulation training games, simulation, operations research, and multi-criteria optimization Page 25.1104.1 c American Society for Engineering Education, 2012 Redesigning a Lean Simulation Game for More Flexibility and Higher Efficiency Abstract Lean manufacturing has become one of the most widely accepted and implemented process improvement methodologies As more companies and organizations embrace lean, effective training and educational programs are in high demand, both in industry and academic programs Due to the practical nature of lean concepts and tools, simulation games have been acknowledged as a very effective pedagogy The interactive contents of simulation games create an active learning environment that enhances the effectiveness of teaching Moreover, with a well designed structure and associated activities, the main concepts of lean, such as flow, visual control, and error proofing, can be demonstrated clearly via a simulation game As a result, various games have been developed by researchers and training institutions in recent years Some simulation games can be conducted in a few minutes (e.g., 5S number game), while some others run for several hours Simulation games designed for classroom purposes, such as the popular LEGO airplane assembly line, typically take to hours to go through several “phases” of lean implementation that demonstrate performance improvement step by step This type of games usually serves the educational purpose adequately However, certain limitations and weaknesses can hinder expected outcome of education For example, most of the games require a fixed number of players, while class size is not always a controllable factor Also, running a twohour game may be difficult for a 75-minute class in university settings Consequently, it is desirable to revisit the design of simulation games to identify improvement opportunities In this paper, an existing lean simulation game designed for office operations is investigated for improvement Using Six Sigma’s DMADV methodology, several problems and limitations of the game have been identified and categorized into Flexibility, Efficiency, Effectiveness, and Cost issues Through cause and effect analysis and several experiments, a revision of the simulation game is proposed with a higher level of modularity and improved standardization and visual aids The redesigned game is able to accommodate a wider range of number of participants, and it can be set up and implemented with less effort The redesign of the simulation game provides a new approach to making lean education a leaner and more productive process Introduction Page 25.1104.2 As lean implementation becomes a standard practice of more companies and organizations, the need for good training programs has increased significantly in the past few years Various workshops and courses for lean training have been developed along with certification programs The use of simulation games appears to be a common trend in lean training due to the practical nature of lean principles Through the semi-practical settings and tasks of lean simulation games, the lean principles can be practiced in a systematic way to demonstrate the benefits of lean implementation step by step As a result, lean simulation games have been very effective in establishing the mindset of lean during the training stage Over the past few years, the authors and their colleagues have used several lean simulation games to facilitate lean training and education programs, including the popular LEGO airplane assembly game1, 5S number game2, lean office simulation game (LOG)3, and some small-scaled computer-based lean simulation games4 Many other lean simulation games can be found in literature as reviewed in the next section With different designs, the games have different purposes and can be applied to different groups of learners, such as college students and industry personnel Among the existing games, the LOG has been used frequently by the authors because the office setting of the game has less resemblance of a factory and can be accepted easily by non-manufacturing personnel The LOG was developed mainly for classroom environment It requires eight participants and approximately two hours to run the game Like many other simulation games, the number of participants of LOG is fixed, and the game activities and required time are not very flexible However, experienced lean trainers would know that the number of participants of a class or a workshop is usually an uncontrollable factor When more than eight people show up, some would be left out of the game, leading to lower effectiveness of training When participants are fewer than the required number, it may not even be possible to run the game as scheduled On the other hand, the two-hour run time of the whole game is too long for a 75-minute class in the university setting Therefore, although it is a very effective training tool, the instructors using this game have been searching for opportunities to improve the design The main objective of this paper is to redesign the lean simulation game, LOG, to make it more flexible and efficient for lean training The target is to accommodate different numbers of participants and different run time as needed, without sacrificing the effectiveness of the performance of lean training The attempt to improve the game was done by using Six Sigma’s Define-Measure-Analyze-Design-Verify (DMADV) methodology, which facilitated a systematic analysis of the problems and root causes to be solved The original version of LOG and the improvement steps are introduced in the following sections, followed by the results and discussions about potential opportunities for further improvement Simulation Games for Lean Training Programs When teaching the practical skills and concepts of lean manufacturing, traditional training methods, such as lecturing and reading, appear to be less effective than using a simulation game Without hands-on practices, the students often have difficulty grasping the concepts or visioning how much impact the lean tools can bring about Consequently, lean simulation games have gained a special role in training workshops and educational programs of lean manufacturing Page 25.1104.3 A lean simulation game typically contains a series of operations that reflect real-world tasks of work In a simulation, several phases of lean implementation are carried out step by step to demonstrate the impact of lean The simplified tasks allow the participants to learn the lean concepts and skills in a reasonably short time Many lean simulation games have been developed in the past Verma5 reviewed 17 popular lean simulation games, such as the TimeWise Simulation of the Lean 101 training program, Aircraft Simulation developed by Lean Aerospace Initiative (LAI), and some games designed by the National Institute of Standards and Technology (NIST) Mirehei et al.3 further extended the list to 34 games and attempted to categorize the impact areas of the games Among the existing simulation games, some of them can be conducted in just a few minutes, such as the 5S number game2 and some smaller-scaled computer-based games4.While some other simulation games can run for several hours, such as the Lean Zone1 which simulates the work flow of a whole office Simulation games designed for classroom purposes, such as the popular LEGO airplane assembly line1, typically take to hours to go through a few phases of lean implementation Interaction among participants is one of the most important components of lean simulation games The interactive environment provides a better learning experience than lecturing or reading In order to facilitate the interactions, however, some common constraints were found in many games, including on-site participation with required number of participants, required setups of game materials and clean ups, and the need of an instructor4 Computer-based games appear to be a good solution to some extent However, based on the authors’ experiences, many existing computer-based lean simulation games have not been able to conduct lean simulation in a full scale as effectively as a physical simulation game For carrying out a classroom-based simulation game successfully, Dukovska-Popovska et al.6 summarized useful guidelines for instructors, including managing the number of participants appropriately, introducing the game before starting, reflections between rounds, and reflections after the game Beside the good practices listed above, the authors also want to take the challenge trying to break some of the common constraints of existing lean simulation games The Original Lean Office Simulation Game before Improvement The simulation game adopted to be improved in this paper is a lean office simulation game, LOG3,7 The game was developed for training personnel involved in non-manufacturing office settings Similar to many other simulation games, the LOG goes through several phases from non-lean environment to lean environment The initial setup of the game is shown in Figure 1, and it requires eight participants to run the game Figure Initial setup of LOG7 In this game, an office of eight employers is in charge of scheduling consulting sessions for customers Two types of customers, represented by red and blue cards in the game, need to go through the office processes of different types of paperwork The availability of the consultant is limited; therefore, the schedulers need to be careful not to overbook the meetings Page 25.1104.4 The game has four phases of lean implementation as shown in Figure Phase I is the initial setting where batch processing is used, and the paperwork and procedure were designed to be error-prone and user-unfriendly Part of the forms to be filled out in the game is shown in Figure The game activities include dice rolling, writing, bubble form filling, signing off, communication for scheduling, etc In the initial phase, typically the game participants would struggle to get only a few customers through the procedure within five minutes of run time In the following phases, lean tools start to kick in, and the performance would improve gradually In the final phase, the process is significantly improved, and the paper form is clear and easy to fill out, representing a “lean” environment The team performance usually improves tremendously from Phase I to Phase IV, including number of customers served, time to complete a service, and number of customers waiting The participants of the game can see the improvements vividly by adopting the lean tools as designed and thus gain the knowledge of lean Phase I Initial Setting Phase II One-Piece Flow (Batch processing, error-prone paperwork) (Reduced queue length) Phase III Poka-Yoke (Overprocessing eliminated, error-proofing) Phase IV Lean Service (Balanced workload, multi-functional workforce, pull system, total quality) Figure Improvement phases and lean concepts involved in LOG … Figure A portion of the customer form used in LOG Page 25.1104.5 From 2009 to 2011, the authors have used the game in numerous training workshops for industry participants and lean-related courses at the university The number of participants in these occasions ranged from one group (approximately eight people) to three groups (approximately 24 people) Feedbacks from the participants were very positive, and several suggestions for improvements were documented and implemented when possible In general, the game is effective although some issues exist As mentioned earlier, the uncertainty of number of participants was the biggest problem The instructors often need to ask a few students to stand by for possible involvement of the game if the participants cannot form one or multiple complete groups Such issues urged the authors to seek further improvement of the design of the game Improvement of LOG: Define Step In this paper, Six Sigma’s DMADV methodology was adapted to help redesign LOG First, the Define step identified the main issues of the original version of LOG and determined the scope of the improvement project The first thing to in Define is to capture the Voice of Customer (VOC) A survey had been collected after every training occasion that involved the use of LOG From the responses of participants, the authors learned that the game is effective, but not all components are exciting Some negative comments from the participants include: • “Filling out too many bubble forms was tedious.” • “The game seemed to be long and started to be boring after a few rounds.” • “The beginning of the game was slow and not very exciting.” On the other hand, the instructors have been exchanging thoughts about the effectiveness of the game and potential improvements There was a consensus that the need to maintain a fixed number of participants to run the game was the biggest burden to the instructors It would be very beneficial if it is possible to accommodate a range of numbers of participants adaptively Also, the game takes approximately two hours to complete, from introduction to conclusion In some occasions, a shorter run time would be necessary, such as a 75-minute or 60-minute class or seminar Some other issues identified by the instructors include the effort to prepare game materials, rigidity of game structure, etc Figure summarizes these issues in four categories of potential improvement of LOG: Flexibility, Efficiency, Effectiveness, and Cost Efficiency Flexibility Total run time can be shorter – Time in each phase can be shorter – Preparation time can be shorter – Number of participants should be more flexible – Total run time should be more flexible – Number of rounds may be reduced – Structure of game should encourage creativity – Game activities can be more realistic – Material costs for each run should be reduced – Other lean tools can be involved – Should be easier to play and learn – Potential Improvement of LOG Manpower for preparation should be reduced – Manpower for gaming should be reduced – Should be more enjoyable – Effectiveness Page 25.1104.6 Figure Opportunities to improve LOG Cost Due to limited timeframe and resources of this improvement project, the most urgent needs for improvement have been identified by a pool of instructors: • Number of participants should be more flexible • Total run time should be more flexible • Total run time can be shorter • Time in each phase can be shorter When the new game was designed, the authors tried to cover the other issues as well Therefore, the scope of the improvement project actually covers more than the priority items Besides, all improvements need to be done in the condition that the educational outcomes would not be compromised Improvement of LOG: Measure Step This step collects data from the original LOG allowing the authors to understand the current performance and problems For the most urgent items (i.e flexibility in number of participants and in run time), related data was found as follows • Total Run Time: The original design requires approximately two hours, including introduction of the game, phases I to IV, pauses for performance review during and after each phase, and final review at the end of the game • Run Time of Each Phase: In the original game, each phase was designed to run for five minutes, plus a two-minute pause at the end of the third minute The total run time of each phase is approximately seven minutes • Number of Phases: The game has four phases as shown in Figure • Number of Groups: The game has been used for occasions ranging from one group (eight participants) to three groups (24 participants) The best results appeared to come from two groups where there were competitions between groups, yet it was not as hard to manage as having three or more groups • Number of People per Group: There were always eight people in a group As stated earlier, the actual number of participants is hard to control Therefore, standby helpers are often asked to fill in some positions in order to form a complete group When extra participants are present, they would be asked to be “observers” without actually playing the game • Performance Metrics: The measures collected in the game include number of customers served, time for the first customer to go through, number of customers waiting in the system, and number of unsuccessful services During the pause at each phase, the number of idle employees and number of customers waiting at each process were also collected Improvement of LOG: Analyze Step In this step, the root causes of the identified issues are determined by analyzing the collected data and current conditions The authors discovered that many of the listed issues are interrelated and may be results of the same root causes Page 25.1104.7 For lack of flexibility in number of participants, the following causes have been identified: • From Phase I to Phase III, the tasks are fixed without any flexibility • • In Phase IV, most positions were turned into multi-functional processors, which provide an opportunity for reducing or increasing the number of participants However, changing the number of players in the last phase would cause more problems All of the tasks were thought to be necessary, which may not be true In response to the above mentioned causes, potential solutions to increase flexibility can be: • Modular design of tasks in the game structure to allow adding or removing participants without affecting the outcome of the game • In a different dimension, it is possible to change the current office setting for two customer types into three customer types to accommodate more players, or to reduce it to one customer type for fewer players For lack of flexibility in run time, the following causes have been identified: • The major problem is the long and fixed run time of each phase in the original design It was found that five minutes of each phase is not really necessary A great portion of the activities (especially in later phases) were to repeat the same tasks (i.e filling the forms) Some test runs showed that as short as minutes per phase is quite sufficient • The pause for review during each phase is beneficial at the beginning two phases But it seems redundant in later phases and does not deliver much information • One of the reasons that the five-minute run time per phase was used is that the bubble forms take a lot of time to be filled If a better form design is used, the turnover rate of forms can be increased and thus eliminate the need to run a long phase • Phase I with batch production and variability introduced by dice rolling often cause lots of confusion Some groups in the past could not even finish servicing one customer within five minutes The activities in this phase may not reflect reality too well With the identified causes, potential solutions to improve flexibility in run time include: • The run time of each phase can be a flexible range from three to five minutes based on the needs and size of class • The pause for review during each phase can be ignored if necessary, especially in the later phases • The forms to be filled out should be redesigned to reduce process time • Phase I may be ignored if necessary With the suggested potential solutions, the next step is to redesign the game and then verify the effectiveness of the new game Improvement of LOG: Design Step Page 25.1104.8 In the Design step, the potential solutions suggested in the previous step were analyzed in detail and tested First, modular tasks were implemented, which allows the chain of the office processes to be longer or shorter, depending on the size of the class Figure shows a few examples of reducing or increasing the number of participants with modular design The key to this design is that Tasks to (Figure 5) are modular and can be used or ignored depending on number of participants These modular tasks are all included in the customer forms to be filled out by players The instructors would inform the participants to use the appropriate sections of the form to facilitate the correct number of participants Due to intellectual property issues, only a portion of the new form is shown in Figure Receptionist & Registrar Task Scheduler Receptionist & Registrar Red Red Blue Blue Receptionist & Registrar Task Receptionist & Registrar Scheduler (a) Players Receptionist Task Task Scheduler Task Task Scheduler (b) Players Registrar Task Task Task Scheduler Registrar Task Task Task Scheduler Red Blue (c) 11 Players Figure Variable number of participants from to 11 players enabled by modular tasks Another key to the improved flexibility of number of participants is that the roles of Receptionist and Registrar can be either combined or separated as shown in (a) and (c) of Figure When the two roles are combined, the game setting can accommodate even numbers of participants Separating the two roles allows for odd numbers of participants to play the game The resulting new design is truly flexible in number of participants, which can be anywhere from six to 11 players in one group When more than 11 players are involved, two groups of the same amount of players or with a difference of one player between groups can be used For example, 14 players can form two seven-player groups; while 15 players can be divided into a seven-player and an eight-player group It is not recommended to have a difference of more than two players between groups, because the dynamics of the groups will be significantly different Some suggested numbers of players for common scenarios are summarized in Table Page 25.1104.9 It should be noted that with longer chain of tasks, it needs longer run time to conduct the game for a meaningful result Therefore, forming a large group is less preferred, if avoidable In addition, the dynamics of combining and separating the roles of Receptionist and Registrar are different Therefore, a delay in receiving customer is added when the two roles are combined (i.e when two people are generating customers) Only one of the two people can be receiving customers at a time Table Recommendations for number of participants in groups Number of Players Players in Each Group 10 11 12 13 14 15 6+1 8+1 10 10+1 6, 6, 6+1 7, 7, 7+1 Note: “+1” refers to separate roles of Receptionist and Registrar Beside the modular tasks, adding or removing one customer type (currently red and blue are used) was identified as a potential solution However, the workload distribution (designed to show the lean concepts) would not be the same any more, and it would change the dynamics of the whole system dramatically Therefore, despite how appealing the solution was, this idea was left as a potential improvement in the future Another significant redesign is the new customer form as shown in Figure The forms for Phase I and II still has the bubbles to be filled out, while the new forms for Phase III and IV use digital numbers that are assumed to be read by computers (in the game setting, not in reality) This new form shows the concepts of standardization of mistake-proofing better than the previous design More importantly, the time to fill out the forms is reduced significantly, which increases the number of customers being served It brings up the morale of the gaming teams, and meanwhile allows shorter run time to show the same effect of lean implementation … Page 25.1104.10 Figure A portion of the redesigned customer forms to be filled out by players In addition, as identified earlier, Phase I in the original design was somewhat unrealistic As an option, this “initial phase” of the game can be skipped when time for the game is limited In a test run, the three other phases still demonstrate the lean concepts fairly well Yet, skipping the first phase, which takes the longest time with learning efforts and frustration, can reduce the game time significantly In order to further eliminate non-value-added activities of the game, the pause for review in each phase is now optional for Phase I and II and not recommended for Phase III or IV Also, as a “side-effect” of the new form with modular tasks, the preparation time for each game has been reduced With the new design, Phases I and II share the same form, and Phase III and IV share another Most of the other preparations (e.g., pencils, erasers, instruction cards, etc.) are one-time tasks and not have to be repeated for each game after the initial game package was developed for the first use Improvement of LOG: Verify Step With the new design, several experiments were carried out to verify the effectiveness For the shortest run time, a full scale test run with two eight-player groups was done with three-minute phases, skipping Phase I, and without a pause within each phase The game was used as a second part of a 75-minute class (after 30 minutes of lecturing) at the beginning of the fall semester of 2011 As a result, the game was completed successfully within 45 minutes It was so successful that, 15 weeks later, students still refer to the game as one of the best activities of the course Several other experiments were conducted to test the new designs for variable numbers of participants ranging from six to nine players From the experiments, the forms and flow of the game went through several iterations of redesign before being finalized Page 25.1104.11 Due to a limited timeframe, not all possible scenarios were tested before the paper was finalized Nonetheless, the authors believe that the results have sufficiently shown that the new design meets not only the urgent needs but also many other potential improvements identified in Figure The results are summarized below • Efficiency: o The total run time can be reduced from two hours to 45 minutes o The time in each phase can be reduced from five minutes to three minutes o Number of rounds can be reduced from four phases to three phases • Effectiveness: o Some unrealistic game activities have been removed o The game is easier to play and more enjoyable, and it is still very effective • Flexibility: o The game can accommodate any number ranging from six to 11 players in a group o The game time can range from 45 minutes to two hours as needed • Cost: o The manpower for running the game was reduced significantly, mainly because the need to have extra standby helpers has been eliminated entirely o The material preparation has been simplified due to forms shared by phases Conclusion This paper presents an improvement project on an existing lean simulation game designed for office operations in order to make it more flexible and efficient Using Six Sigma’s DMADV methodology, several problems and limitations of the game have been identified, followed by potential solutions to resolve the problems Through iterations of experiments and redesign, a new simulation game has been created that is flexible enough to accommodate variable numbers of participants and variable length of run time as needed Consequently, the new game is more flexible and efficient as expected Although in the mindset of lean, there is always room for more improvement The authors have been improving the game continuously in the past and will continue improving it in the future Several issues of the original game identified in the paper appear to be common for many lean simulation games, such as the fixed number of participants Therefore, the improvements suggested in this paper can provide some directions for potential improvement of lean simulation games in general Furthermore, several other ideas noted in the improvement project were not implemented yet For example, the additional customer type can be a good option if designed well, and changing layouts during the game may create opportunities to showcase more lean principles and systems thinking These options should be further explored to potentially create better simulation games for the education of lean manufacturing in the future Bibliography Visionary Products, “Lean zone office,” (http://www.visionaryproducts.biz/Portals/0/Files/ VPF-0001_Visionary-Products.pdf) Superteams, “The 5S Numbers Game,” (http://www.superteams.com/5s-game.php) Mirehei, S M., Kuriger, G., Wan, H., and Chen, F.F., 2011, “Enhancing Lean Training for the Office Environment through Simulation and Gaming,” International Journal of Learning and Intellectual Capital, 8(2), pp.206-221 Wan, H., Tadikonda, B.M., and Kuriger, G., 2011, “Lean Training via the Internet: Two Flash-based Simulation Games,” 2011 Annual Industrial Engineering Research Conference, May 21-25, Reno, NV Verma, A K., 2003, “Simulation Tools and Training Programs in Lean Manufacturing – Current Status,” A Technical Report submitted to NSRP-ASE, National Shipbuilding Research Program Dukovska-Popovska, I., Hove-Madsen, V., Nielsen, K B., 2008, “Teaching Lean Thinking Through Game: Some Challenges,” Proceedings of 36th European Society for Engineering Education (SEFI) on Quality Assessment Mirehei, S M., 2009, “Design and Development of a Simulation Game for Lean Administrative Process Training,” Master Thesis, The university of Texas at San Antonio, UMI No 1464009 Page 25.1104.12