Printer-friendly version

Simulation Game for Peach Retail Ordering Systems
Deepak Aggarwal, University of Georgia
Stanley E. Prussia, University of Georgia
Wojciech J. Florkowski, University of Georgia
Don Lynd, AMS, United States Dept. of Agriculture

Abstract
Despite the progress in the development of simulation games and models for training in many areas, the post harvest produce sector seldom has been a subject of instructional tools. A computer-mediated simulation game for this objective was developed in the Windows environment. It uses Stella software (www.iseesystems.com) to model the system and to develop the simulation equations. The game, developed with the complicated needs of the produce sector in mind, combines systems dynamics and systems thinking to model and simulate the essential elements responsible for meeting consumer demand. The player assumes the role of a retailer, striving to minimize inventory without running out of an item. The player selects a consumer demand from various deterministic and stochastic consumer demand options. Players are challenged by delays in produce delivery (two weeks) and perishability of the produce. Tension is caused by conflicting objectives, such as running out of produce or avoiding the penalty resulting from excess inventory. If required, the player can reduce his excess inventory by exercising his option of putting produce on "sale." The bank balance shows the impact of the decisions made. Players are surprised by the difficulties encountered for maintaining low inventories while maintaining an uninterrupted supply, even at the easiest level. The simulation emphasizes the importance of produce ordering based upon deterministic and stochastic portrayals of consumer demand, the delays in produce delivery by the system, perishability, and a measure of financial assessment of the retailer's performance. The game provides a participatory, hands-on experience for understanding the concepts and for developing skills desired in post harvest fresh produce businesses. The game was used as a tutorial for an undergraduate university course on food science and at three international symposiums. It was also played by high school students and by a produce manager at a major grocery. It can be used as a teaching tool by marketing, postharvest engineering and biology students and as a training tool by produce department personnel in retail outlets to understand the dynamics of supply chains in the post harvest marketing system.

About the authors...

1. Introduction
Decisions by produce managers determine the quality and availability of fresh produce at retail stores. Fresh produce enters a supply chain where it is grown and flows through various links to retailers and on to consumers. Quality is lowered when excess product is ordered and held at retail for several days before being sold. Equally critical are empty shelves that result if insufficient quantities are ordered. Produce managers at retail stores would benefit from continuing education that increases their understanding of supply-chain system concepts. Likewise, students need to learn practical aspects of supply chain management. A result of a training tool would be enhanced understanding of the needed ordering skills for increasing the probability of uninterrupted supply, minimum losses, and increased profits.

The "teaching by telling" system of training is history. Today's employees, raised on television, personal computers and instant microwave foods, must be intrigued and actively motivated with attention-getting "learning-by-doing" training programs (Schank, 1997). Schank further stated that effective training could be imparted through the virtual reality of computer simulations and role playing scenarios. Simulation gaming mimics a real process through a user interactive game. Players learn and improve their knowledge about the intricacies of a process. User interactive games create learning playgrounds based on scientific principles. Games simulate the behavior of a process or a system in a simplified environment. The simplification of complex problems enables the players to gain insights about the essential elements of the system or stages of the process. Simulation games are established methods for improving business management, industrial design, logistics, and military training. They allow for events that would be too dangerous and costly to attempt in reality. For similar reasons, pilots use flight simulators to gain skills that would be needed during extreme flying conditions and situations.

A major agricultural policy report released by USDA (2001) explicitly stated the need for consumers to drive agricultural systems. Managers of all links in produce supply chains have not recognized the central role of consumers within the system. Although, food retailers are aware of consumer buying behavior, more often than not, they blame consumer buying variations as the major cause of losses in produce retail. Also because of its perishability, fresh produce is a category that is difficult to manage. Improved retailing decision-making skills can increase profit margins in produce marketing. The existing teaching tools in the produce industry do not include management simulation games similar to those used in other industries. Retailers list employee-training needs among top priorities. Similarly, instructors at colleges and universities lack tools that realistically portray fresh produce retailing that could sensitize students to the subtleties of managing economic and postharvest engineering aspects of produce handling. A simulation game will increase the awareness of students, retailers, and others within agricultural systems about the interdependencies and relationships among various links. Recently, Prussia (2000) outlined the need and requirements of a simulation game as a training tool for the retailing link of fresh produce supply chains.

The objective of the project presented in this paper was to develop a computer game that simulated the ordering of fresh produce by retailers. Specifically, the game should simulate some of the critical elements of postharvest supply chains for fresh produce as faced by managers of produce departments in retail outlets. The game should meet the objective using uncomplicated rules within a simple, but realistic situation. The game should be well-paced and interactive to engage players.

1.1 Review of Literature
Sterman (2000) pioneered the use of games in teaching the complexities of supply chains. His board management game based on systems dynamics, known as "The Beer Game," has been in use for about three decades. Many games have been developed with the advent and increased use of personal computers. This technological advancement did not change the purpose of games used as teaching tools, but has allowed for increased flexibility by incorporating an increasing number of elements from simulated real life systems. Shewfelt and Prussia (1993) provided the concept of systems approach to post harvest handling of fresh produce.

Hofstede et al. (2002) developed a management game where participants establish a simulated international food industry chain. The game was designed to present a number of chain problems that occur in an international food supply chain. Participants process various fruits and the end product of the chain is a fruit pastry sold and consumed by participants. Schotzko and Hinson (2000) state that it is possible to remove time and cost from supply chains, improving profitability through conceptual advances and use of computer hardware and software. Duke and Malcon (2003) use game theory to develop an economic model of producer and residential-neighbor behavior for investigating the ways producers balance the choice of management practices with the risk of agricultural-nuisance lawsuits. Smith (2001) compared different levels of interaction in a computer game-like situation as a means of learning efficiency on an internet-based spatial visualization task, involving polyomino puzzles. He concluded that alternating between interaction and observation is the best way to learn spatial visualization.

Hwang and Esquembre (2003) developed "easy Java simulations," an interactive tool for the conceptual learning of science. Jong and Sarti (1994) discuss the development of new models, methods, and tools to support the design and production of (computer-based) learning material. They describe a number of projects that represent the state of the art of European research on courseware authoring, with a special attention to simulation-based learning material. Sawhney and Mund (1998) reviewed the role of simulation education in "construction management" and the need for developing simulations for it. Johns (1999) demonstrated the use of web based practice environments as a tool to teach mechanical skills.

However there are no available games in the field of supply chain dynamics in fresh produce, especially for the retailers.

2. Methodology
The game we developed for simulating retailing of fresh produce is based on systems thinking. Senege et al. (1994) state that "at its broadest level, systems thinking encompasses a large and fairly amorphous body of methods, tools and principles, all oriented to looking at the inter-relatedness of forces and seeing them as part of a common process." Systems Dynamics, developed by Forrester (1961), has become particularly valuable as a language for describing how to achieve a fruitful change; it allows one to construct a model in which the "state" of the system changes with the rate of input and output of each variable that can be monitored (Williams et al. 1995).

The human component is crucial to understanding why a desired outcome may or may not be achieved; thus, human behavior plays a major role in the business decision-making process. By combining systems thinking and human behavior, soft systems methodology (Checkland and Scholes, 1990) can be used to create a simulation game that requires players to integrate systems thinking and system dynamics into their decisions.

A visual interactive package, Stella 7. 0, was employed (ISEE Systems, formerly High Performance Systems, Inc., Lebanon, NH, USA, www.iseesystems.com) to develop a simulation game for produce retailing. Stella is an icon-based systems thinking modeling software that enables the overall performance of the system to be visualized and simulated. Stella 7.0 is available for Microsoft Windows operating systems. The basic elements of the Stella language are stocks, flows, converters, and connectors. From an operational point of view "stocks" are the accumulators and indicate the total amount that they are accumulating over time. The "flows" are used to describe activities or changes causing modification in the "stocks." The "converters" modify the activities within the system, but, unlike the stocks, they do not represent an accumulation of anything, nor do they have a memory. The "connectors," as the name suggests, are used to connect the stocks to the converters, the stocks to the flows, flows to each other, converters to flows, and converters to converters. Aggarwal et al. (2003) provides more details on how Stella was used for developing a simulation that functions as a game.

An external link to Stella Software.

3. The Game
The game allows for different levels of complexity. Starting from the "introductory level," options can be selected to include additional parameters and relationships. Common to all selections are four basic underlying sectors: namely peach consumer, retailer, wholesaler, and farmer. The central role of variable consumer demand in the peach supply chain dictated the selection of the dynamics of the consumer-retailer interaction as the beginning stage. The simulation was named "the peach game" because it uses peaches as the primary product, but the concepts would work for any fresh fruit or vegetable. Two players can alternatively play the game, each trying to outwit the other in the same produce market.

The purpose of the "introductory level" is to initiate the player to the game and the simulated system dynamics. The player assumes the role of a retailer ordering and selling peaches. The roles of the consumer, wholesaler, and farmer are assigned to the software program. The player (retailer) strives to satisfy consumer demand. The player faces the predetermined pattern of quantity demanded over a period of weeks and is aware of the two-week delay in the arrival of the ordered quantity of peaches. The mission of the retailer, as presented to the player, is to make sure peaches are available at all times, yet to minimize the fruit inventory. The game score is the cumulative excess inventory expressed in dollar value (bank balance). Because the value of excess inventories represents a loss to the retailer, the player's aim is to minimize it without running out of stock.

The "quantity demanded" that the player can expect in the coming time periods (weeks), represents historical records. Players can view the average weekly quantity sold for each week in the past. Figure 1 (a) shows the graphic change in the quantity demanded over time for the introductory level: the quantity, measured in the number of boxes sold, increases as the harvest progresses, peaks in the middle of the season, and subsequently declines. Plotted quantities create a linear pattern with two sub-periods of a steady increase and a steady decrease in the quantity demanded.

Players can change orders each week. The volume sold last week and the current inventory is plotted instantly and can be visually assessed. However, the current inventory reflects past orders. The graph provides relevant information for ordering additional deliveries subject to the two-week delay. The game ends if a player is unable to meet the demand in a particular week. The message posted on the screen informs the player about the end of the game. The simulation resets itself and a new game can begin. Thus, a tension exists between minimizing inventory and running out of peaches.

The purpose of the introductory level was to make the game as simple as possible. A step function was first used for "consumer quantity demanded." However, tests showed that a ramp function enabled players more time to use feedback to make corrections. Despite its simplicity and the known quantity demanded in the introduction level, the two-week delay of deliveries makes it difficult to keep a balance between maintaining peaches in stock and minimizing inventory.

3.1 Changing Quantity Demanded
After players master the "introductory level" they can move to more difficult scenarios. In these, the player has a number of different demand stream options, such as a normal distribution (Figure 1b). Another option is a stochastic demand having a mean of 10 cases per month with a standard deviation of 2 (Figure 1c). The fourth option is also a stochastic demand, but it is "randomized normal distribution" (Figure 1d). The stochastic demands make playing the game more difficult than the levels having deterministic demands. The player also has an option to generate any desired consumer demand history and then to simulate the effect of ordering decisions on the game outcomes. In reality, consumer demand is unpredictable. In one simulation option, players must plan orders for an up-coming peach festival only to have a hailstorm hit the peach orchard, causing reduced supplies.

3.2 Adding financial elements
The player has a bank account, which can be set to a selected value when the game is initialized. It decreases when the orders are placed and increases when the revenues are generated from selling the product. The profit on each case of the product sold is 50%. There is an option of putting the peaches on sale. The sale doubles the demand but the profit margins decrease to 20% from the usual 50% . There is a dollar penalty for excess inventory, which is 5% of the price of the product each week. The bank balance is plotted as a bar chart on the play page of the game. The score of the game is the bank balance that the player strives to maximize.

3.3 Perishability
All the above options apply for retailing any product and not fresh produce in particular. A major problem faced by the retailers in fresh produce marketing is quality deterioration, which results from physiological senescence. So an element of perishability was introduced which can be selected by the player. If a delivered order cannot be sold within two weeks its quality deteriorates to the point that it must be discarded.

The critical elements in the game are the quantity demanded distributions, two-week delay in order arrival, bank balance, ordering, sales, and perishability of the fresh produce. If the player has ordered in excess of the consumer demand, then the excess inventory can be managed by putting the fresh produce on "sale" and preventing the produce from perishing.

Figure 2 is a screen shot showing the options described above. A screen shot of one of the game outputs is shown in Figure 3. The area below the inventory curve signifies the cumulative excess inventory, which must be minimized to obtain the best score (bank balance). However, if the inventory curve goes below zero, the simulation terminates itself, signifying that the player has not been able to satisfy consumer demand. If the inventory increases beyond a set value, the player gets a message with a suggestion to reduce the inventory.

Figure 1 a, b, c, d. Various consumer quantity demanded streams (cases/week)

Full size image.

Figure 2.Screen shot of options page of the game.

Full size image.

Figure 3.Screen shot of simulation page of the “Peach game."

Full size image.

The model was validated using a number of scenarios, performing the computations manually and comparing the results of the manual computations and the model output.

4. Feedback 
As the game was being developed, various players, including selected faculty and staff at the University of Georgia, the head of the produce department of a retail store (two versions of the game), and about 100 students provided feedback which resulted in three iterations or versions of the game. For example, the produce manager of the retail store suggested the inclusion of dollars as the instrument for measuring players' performance instead of a score (the measure of performance in an earlier version of the game was the area under the inventory curve).

The developed game was used as a tutorial for an undergraduate course in the Department of Food Science at the University of Georgia. The students downloaded the game and the required software to play the game from the web on their computers. After playing the game by themselves, they provided answers to a questionnaire based upon the game. Not only were they asked about what they learned from the game, but also they were required to evaluate it. Table 1 gives a few examples of their responses to some of the questions. Table 2 and Table 3 summarize the responses of 40 students for the "likeness" and "educational" attributes respectively. The weighted mean of 4.125 for "likeness" attribute signifies that the students evaluated the game as being between "liked very much" and "liked moderately." Similarly the weighted mean of 4.15 for "educational" attribute signifies the educational value to be between "highly educational" and "moderately educational." The game was also played by high school students in two consecutive years. About 80 students were subdivided into four groups. Each group formed two teams and each team competed to outwit the other. The game was demonstrated at three international conferences in 2002 and 2003. The conferences were in the areas of post harvest technology, food distribution, and modeling.

5. Debriefing and Discussion
The game can be followed by a debriefing session to review the complexities of fresh produce retailing. The lessons taught by the game emphasize several issues in fresh produce supply chains: 1) the stochastic nature of the consumer demand as measured by the quantity consumed in each time period, 2) the need to minimize current inventories, 3) the effects of the delays in the delivery inherent in the system, 4) perishability of fresh produce, and 5) the placement of produce on sale as a tool for managing inventories and meeting demand. An instructor can place the relative emphasis on all or any of the specific areas. The selection will determine whether the game will take one or more continuing education sessions or class periods to complete.

For college instruction the developed produce delivery game emphasizes the interconnectedness of implications stemming from a single decision, such as ordering a shipment of fresh peaches. The systems approach to delivery is illustrated in the multitude of relationships sensitizing players to the multidisciplinary nature of produce handling. Initially, the players typically order too much and as they reduce subsequent volumes ordered they are unable to meet the consumer demand. After the players become accustomed to the game dynamics, they learn the underlying concepts. Foremost, game users recognize the uncertainty of the produce delivery system. The degree of uncertainty affects differently various links in the system, but it all begins with the discrepancy between the anticipated and the actual consumer behavior. This results in a bullwhip effect. The game is consumer-driven, although the option to declare an item as "on sale" provides some short-term flexibility to retailers. By assuming the role of a retailer, farmers could recognize the difficulties in ordering produce. Intermediate links in the system are sometimes accused of taking an unfair advantage, but retailers may commit ordering errors even when having full access to information about past sales.

6. Conclusion
The highly competitive sector of produce delivery systems is driven by consumer behavior. Despite the progress in the development of simulation games and models for training in many areas, the produce sector seldom has been a subject of instructional tools. The game, developed with the complicated needs of the produce sector in mind, combines systems dynamics and systems thinking to model the essential elements responsible for meeting consumer demand. Specifically, the game requires the input of human decisions in systems dynamics and thinking during the simulation. It is designed to help retailers improve fresh produce ordering decisions without risking the actual consequences of sub-optimal decisions or unpredictable outcomes in an actual market environment. The game provides a participatory, hands-on experience that allows players to understand the concepts and develop the skills desired in the fresh produce industry.

7. Acknowledgement
The funds for the project were provided by Marketing Services Branch, Agricultural Marketing Service, United States Department of Agriculture, Washington, USA. The help of Dr. Robert Shewfelt, Professor at UGA, in letting the Food Science and Technology students play the game is thankfully acknowledged. The peach game (version 3.1) can be downloaded from the UGA Biological and Agricultural Engineering website at www.griffin.uga.edu/ageng/programs/peachgame.exe
The execution of the game also requires a freely downloadable save disabled ‘demo’ version of ‘Stella’ from www.iseesystems.com.

An external link to peach game.

8. References

Aggarwal, D, Prussia, S.E., Florkowski, W.J., and Lynd, Don. (2003). Simulation model for fresh produce supply chains. ASAE Technical paper number 03-6211, St. Joseph, MI.: ASAE.

Checkland, P., Scholes, J., (1990). Soft systems sethodology in action. Chichester, England: John Wiley & Sons.

Duke, J.M., Malcon, S. A., (2003). Legal risk in agriculture: right-to-farm laws and institutional change. Agric. Syst., 75 (2-3), 295-303.

Forrester, J. W. (1961). Industrial dynamics. Cambridge: MIT Press.

Hofstede, G. J., Trienekens J. H., Ziggers G.W. (2002). The strawberry game. Online. http://www.info.wau.nl/people/gertjan/straw. html. [Retrieved 25th Feb 2004].

Hwang, F and Esquembre, F. Easy Java simulations: An interactive science learning tool. IMEJ, 5 (2). http://imej.wfu.edu/articles/2003/2/01/index.asp

Johns, J. F. (1999). Web-based practice environments to teach mechanical skills. IMEJ, 1 (1). http://imej.wfu.edu/articles/1999/1/01/index.asp

Jong, T.d. and Sarti, L. Editors (1994). Design and production of multimedia and simulation based learning material. Dordrecht, Kluwer Academic Publishers

Prussia, S. E. (2000). Soft systems methodologies for modeling postharvest chains. Acta Hort. 536, 653-660.

Sawhney, A and Mund, A. Simulation based construction management learning system. In D.J. Medeiros, E.F. Watson, J.S. Carson and M.S. Manivannan, eds. Proceedings of the 1998 Winter Simulation Conference.
http://www.informs-cs.org/wsc98papers/179.PDF

Senge, P. M., Kleiner A., Charlotte R., Ross R.B., Smith B.J. (1994). The fifth discipline fieldbook -- Strategies and tools for building a learning organization. New York: Doubleday.

Schank, Roger. (1997). Virtual learning--A revolutionary approach to building a highly skilled workforce. New York: McGraw-Hill.

Schotzko, R.T., Hinson R.A. (2000). Supply chain management of perishables: A produce application. J Food Distribution Research, Vol XXX1 (2), 17: 25.

Smith, G.G. (2001). Interaction evokes reflection: Learning efficiency in spatial visualization. IMEJ, 3 (2). (http://imej.wfu.edu/articles/2001/2/05/index.asp)

Shewfelt, R.L., Prussia S.E. (Editors) (1993). Postharvest handling: A systems approach. Orlando, FL: Academic Press.

Sterman, J. (2000) Business dynamics: Systems thinking for a complex world. New York: Irwin/McGraw-Hill.

USDA, (2001). Food and agricultural policy--Taking stock in the new century, Secretary of Agriculture Report. Online. http://www.usda.gov/news/pubs/farmpolicy01/fpindex.htm [Retrieved 25th Feb 2004]

Williams, T., Eden C., Ackerman F., Tait A. (1995). The effects of design changes and delays on project costs. J. Oper. Res. Soc., 46(7), 809-818.

********** End of Document **********

IMEJ multimedia team member assigned to this paper	Yue-Ling Wong

© 2004 Wake Forest University (from Volume 6, Number 1, of The Interactive Multimedia Electronic Journal of Computer-Enhanced Learning).