Simulating Print Production System through System Dynamics Modeling towards Productivity Improvement in Order Fulfillment Ma. Crea Eurice D. Santos Abstract: The production system in logistic operations is an important function that significantly determines an organization’s performance in delivering customer requirements. This study investigates on the production system of a generic commercial printing company with the goal of understanding interrelationships of operation parameters towards influencing print productivity in relation to business logistic function of order fulfillment.
Simulating the cause and effect relationship of print production work flow considering feedback mechanisms presents a comprehensive representation of the system’s complexity whereby interplay of associated variables and parameters are examined. Specifically, through developing alternative mental models, opportunities at both strategic and operational level can be exploited based on impact assessment of policies and decisions from simulation runs of different scenarios Keywords: System Dynamics, Print Production, Order Fulfillment 1. Introduction
Worldwide manufacturing industries are in constant search for means of producing faster, cheaper, and better products or services(Uribe, 2008). In line with this, production takes a dominating and strategic logistic function in the process of order fulfillment (Helbing & Reichel, 1998). However, despite increasing technology in the printing industry added to widely available management tools, print productivity still has fallen further behind other manufacturing industries (Grant, 2003; Uribe, 2008; Zeng, Lin, Hoarau, & Dispoto, 2009).
In response to this, a system dynamics approach, modeling print production system, is provided to better understand the interrelationships between operation parameters and to uncover feedback loops influencing overall organization productivity. Analysis of the cause and effect relationships of print production variables exposes underlying reasons for lagging behind the average productivity levels. Specifically, simulating print production mental models provides an overview of system behavior whereby opportunities at both strategic and operational level are identified and exploited.
Commercial printing embodies a diverse, complex, and dynamic system(Uribe, 2008; Romano, Fawcett, & Soom, 2003). Its diversity is rooted to wide customized product offering possibilities. Particularly, commercial printers are involved in the manufacture of various custom printed products including flyers, posters, brochures, books, newsletters, invitations, packaging materials, and many more(Romano, Fawcett, & Soom, 2003). This make-to-order production system makes productivity improvements quite a challenge(Uribe, 2008).
Industry complexity, on the other end, transpires as diversity is complemented with mass production. More to this high customization high volume not-so-good match structure, complexity is intensified as it is classified as both a capital- and labor- intensive industry (Uribe, 2008; Zeng, Lin, Hoarau, & Dispoto, 2009). The highly diverse and dynamic job mix hence result to different print system configuration involving several work flows and varying combinations of equipment, resources, and business models (Kipphan, 2001; Uribe, 2008).
In view of that, this highly complex and dynamic nature of print production system makes it viable for a system dynamic approach investigation. In so applying systems thinking, further knowledge and comprehension enables management to not only create better strategies and decisions but more so to determine the effectiveness of the proposed solution immediately, without the need of waiting for results and actual application to the system which may be too costly and risky.
Generally, there is much to explore in the world of printing. Today, just like several other industries, it has been highly populated with printers that can produce good quality printing, and what differentiates a printer from another is basically a combination of the added value and services it is able to provide (Dana). Thousands of different mix and match of strategies have been designed and employed by commercial printers yet no single one is deemed the best and only solution (Uribe, 2008).
Assessment of various commercial printers’ current situation and problems point to disconnected or misaligned process requirements in producing print materials, whereby disintegrated equipment or process efficiencies are accounted distinctly. Adopting a system dynamics approach of exploration, focusing on the relationships interconnecting the complex system of printing rather than on individual areas of production, provides a holistic view whereby opportunities for improvement and better decision-making can be built. 2. Problem Articulation
Problem Statement and Research Objectives Productivity has been widely acknowledged as among any organization’s critical success factors. However, evaluating and understanding this measure entails a rather complex process(Kipphan, 2001). Beginning with the question of how to measure, Millet and Rosenberg, as cited in Uribe (2008), characterize productivity as the relation of throughput with a given set of resources, thus translating the ability to produce higher output with a fixed set of resources to be classified as more productive.
In this study, throughput through time is the accounted measure of productivity as per production backlog and turn-time estimates. Research indicate that the behavior of in-process production follows a cyclical pattern, increasing and decreasing accordingly with demand and turn time. Figure 1 shows the reference mode for the production backlog based on available literature. Figure 1: WIP in an Uncontrolled Process (Pritschow & Wiendahl, 1995) The high complexity of print production systems outlays difficulties in realizing high productivity which in turn contribute to unsatisfactory order fulfillment performance.
Being influenced by various interacting factors that form feedback mechanisms, straightforward linear solutions may not automatically result to observable higher productivity. Apparently, most decisions in printing companies are subject to miscalculations taking its roots from managers’ personal assessment on the basis of individual expertise. Basically, decision foundation builds on trial-and-error solution implementation originating from manager’ experiences(Zeng, Lin, Hoarau, & Dispoto, 2009).
Accordingly, the higher the complex¬ity of the system, the more likely decisions are to be errant. Incidentally, printing managers’ idea of a “perfect print shop,” can be highly diverse and conflict¬ing(Uribe, 2008). Hence, in order to address the rather low productivity evident in the printing industry, a holistic perspective analyzing the design and management of print production as an integrated system is simulated using a system dynamics approach to model the causation among interrelated variables and identify leverage points for designing organizational policies and decisions.
Generally, the modeling of print productivity in terms of system dynamics allow for a comprehensible depiction of complex relationships, thus subjecting underlying assumptions and relationships expressed in terms of functions and differential equations to be evaluated over time and scrutinized(Richmond, 2001). With this in mind, specific objectives of the study are identified as follows: create a computer simulation model that conceptualizes the dynamics of a generic commercial printing production system •to provide an understanding of the current state of commercial printers so as to indicate a path for higher productivity •identify policies and decisions towards opportunities for improvement through simulation runs of different strategy scenarios Study Scoping In printing job orders, process is quite fascinating, having to involve high-speed machines, reams of paper, metal plates, and many other supplies.
Also, service may begin from layout down to the binding process. An overview of the entire printing process is illustrated in Figure 2. Figure 2: Overview of the Printing Process Discussion of the printing process in detail can be divided into five main phases involving coordination between consumers and service providers (Refer to Figure 3). Figure 3: Printing Business Process Map (Glykas, 2004) In the order acquisition phase, commercial printers receive an order from a customer and coordinate to determine job order specifications and project concept.
The following phase, consisting of the design stage, may be a joint venture for print buyers and providers whereby the design team prepare basic design for the print job based on the discussed concept. In the electronic production phase, all scanning, editing and electronic image processing is performed ensuring compatibility with printing mechanism. The customer shall then signal the printer to proceed with the printing of the project whereby film production is initiated if CTP technology is absent to the provider.
In this phase, films produced are put in the proper layout, and then blueprints and digital proofs are sent to customers for approval. Upon approval from the customer, the printing phase is commenced whereby negatives are converted to plates to be used in the offset printing machine. The machine is set up with the appropriate colors and paper and printing impression is performed. From printing, printed sheets undergo a finishing phase whereby all cutting and binding activities are performed.
Finally, the delivery phase includes all activities related to final inspection, packaging and delivery (Glykas, 2004; Zeng, Lin, Hoarau, & Dispoto, 2009). However, the scope of the study shall only include the film production, printing, and finishing phase of the production system. 3. Dynamic Hypotheses and Model Presentation A mental model of a generic print production system is developed with causation hypotheses adopted from available researches and knowledgeable people from the industry. Figure 4 illustrates the causal loop diagram developed to describe the interrelationships between variables and associated feedback structures.
Again, the purpose of the model is to provide an understanding of the current state of commercial printers and to create an avenue for higher productivity. Further explanations of the variable interrelationships in relation to this are discussed in the following section as per stock flow diagram. Figure 4: Print Productivity CLD The mental model incorporated the interaction between sales variables, production backlog, customer satisfaction, and production capacity whereby sales orders are triggered by relative price, sales aggressiveness and customer satisfaction which in turn adds up to in-process production stock delivered as per capacity.
This time delay of fulfilling an order constrained by capacity influences customer satisfaction level specifically modeled by the effect of too much production backlog towards turn time and quality. Production Stock Flow The production stock is modeled as per the three basic production phases in offset printing, namely, the (1) pre-press, (2) press, and (3) finishing or bindery (Mine). Pre-press covers the stage wherein digital files are converted into film negatives to be used in creating metal plates for press run. Press includes the actual impressions of images or texts to papers.
Lastly, Postpress or Finishing comprises the binding of printed material or any finishing services that may be required of the job specification until it is packaged and made ready for delivery (Introduction to the Offset Printing Process). Print production operates under the assumption of a make-to-order policy. Hence, stock and flow struc¬ture are represented by orders coming into the production system, processed for a given time (production delay or turn time), then shipped to customers(Uribe, 2008). The production backlog represents the in-process orders waiting to be completed — Prepress + Press + Postpress.
It increases as sales order exceeds the output capability of the system and vice versa. It flows from the Prepress phase as plates are processed constrained by associated capacity and influenced by pressure or processing factor based on perceived WIP level. These processed plates then serve as an input to the Press stage. In the Press stage, there is an added outflow and inflow based on print rejects and returns that is unacceptable to standard quality and require reruns respectively. The Postpress stage likewise follows the same underlying mechanism with an option to outsource finishing requirements.
Figure 5: Production Stock Flow The rest of the model concentrates on the feedback structures that control the system. Order Inflow The overall production stock is affected by the rate of the order inflow coming in to the system at the prepress phase generated by sales orders from customers (Uribe, 2008). The assumption is that all orders undergoes the three print production stages as is often the case in most commercial printing services. In modeling production systems, it is very important to understand what affects the production stock including the inflow and how this rate can be managed.
Generally, order inflow can be affected by multiple variables. Refer to Figure 6 for significant factors considered by consumers in print service provider selection. However, inputs to the model focus on identified variables from literature as relevant to the problem and to the policies to be designed in the simulation. Figure 6: Factors Considered in Selecting a Printing Services Provider(Pellow, 2003) The first variable accounted in the model having direct impact on the sales order rate is the relative price, defined as the ratio of the price of the company to the market price.
Demand curve indicates that demand increases as price decreases. On the basis of this basic demand theory, if relative price is less than one, meaning price offered is less than market average, then the order rate increases, and vice versa. The second variable considered is the average order. To replace the complexity of the market demand, this is modeled as a constant indicating that a printing company that has been established in a market will have an average inflow of orders based on historical order patterns. The third variable is described as a switch variable on sales aggressiveness.
Lastly, customer satisfaction level is also considered to account for the top three ranking factors in printer selection: 1) Dependability, 2) Quality, and 3) Turn Time. Many more variables could be included in the model, yet current variables selected are assumed adequate for the model purpose and boundaries. Figure 7: Order Inflow Customer Satisfaction Level Customer satisfaction level generally creates the feedback mechanisms within the production system as per the effect of high level of satisfaction towards increasing sales order that likewise increases production backlog.
In particular, increasing backlog may be exceeding production capacity thereby increasing turn time as well as probability of rejects. This in turn decreases customer satisfaction which then creates the balancing loop. Fundamentally, the main reasons why print buyers abandon printers are due to quality issues and late deliveries (Merit, 1992; Uribe, 2008). However, the behavior of customer satisfaction is observed to follow the trend of turn time considerably more to quality as per simulation study. Figure 8: Customer Satisfaction Level
Figure 9 illustrates the integrated stock flow diagram of the print production system. (Refer to Appendix for associated equations) Figure 9: Print Production SFD 4. Simulation Model The behavior of system is simulated and compared with reference mode shown in Figure 1. Using Stella, resulting behavioral pattern is presented in Figure 10. Figure 10: Simulated WIP Behavioral Pattern The behavior of the production backlog follows the oscillating trend of the reference mode as defined by sales order and shipment rates.
Apparently, in-process level is cyclical in nature, depicting the principle of limits to growth. Figure 11: Reference Mode Figure 12: Simulation Behavior Results indicate that production backlog follows the trend of customer satisfaction with increasing orders as satisfaction rises more to other variables considered in the study. However, it is likewise observed that a limit to the growth of sales orders that serves as input to production backlog exist given that a printer can only accommodate orders pertinent to their capacity.
As capacity represent the controlling parameter of production outflow, increasing backlog is accumulated when demand exceeds capacity thereby increasing turn time and reducing customer satisfaction. This balances the upsurge in backlog as sales order is limited, thus creating the cyclical pattern. Figure 13: In-Process (Current System) Figure 14: Performance (Current System) In order to evaluate effectiveness of alternative proposals, turn-time, backlog, and total orders are tracked to compare system performance with alternative scenario simulation runs.
In line with this, current system accommodated orders as well as production backlog and turn time range is presented in Table 1. Table 1: Performance Measures VariableMeasures Total Orders164. 07 Production Backlog2. 91 to 23. 92 Turn Time1. 164 to 7. 974 Sensitivity Analysis In order to better understand the behavior of the system as per management decisions, changes in the parameters are subjected to sensitivity analysis. Variables chosen for assessment were considered owing to apparently significant contribution towards the behavior pattern of the system.
In line with this, system is observed to be sensitive to capacity parameter. Hence, effect of prepress, press, and postpress capacity changes to turn time, backlog, and total orders behaviors are examined specifically. Graphs in Figure 15, Figure 16 and Figure 17 depict the results of increasing and decreasing system capacity (1- decrease, 2- sustain, 3-increase). Figure 15: Turn Time Sensitivity to Increasing Capacity Figure 16: Backlog Sensitivity to Increasing Capacity Figure 17: Total Orders Accommodated to Increasing Capacity
Results indicate a logical behavior depicting decreasing turn time and backlog while increasing total orders as capacity is increased. Turn time behavior is observed to have high range for a decrease in capacity. On the other hand, increasing capacity produces more fluctuating results with average turn time slightly lower than current capacity. Nevertheless, this is still considered a better option given that turn time variance is seemingly lower despite shorter time frame of increase and decrease of values. Specifically, it depicts a rather more consistent order fulfillment lead time to end-consumers.
Alternative Scenarios Increasing Factor of Sales Aggressiveness The first alternative scenario modeled is to increase the factor of sales aggressiveness to managing sales order based on production backlog. In particular a policy on pushing sales account executives when work in process is below average backlog to smooth in-process and avoid idle capacity is proposed. However, results indicate no significant change in behavioral pattern as shown in Figure 18 and Figure 19 with the exception of shorter variation in sales order inflow.
Effects on range of performances reflect minimal improvement as shown in Table 2. Nevertheless, this can still be considered a better alternative with slight increase in performance without incurring additional costs for the provider other than having better demand management with the help of current workforce. Figure 18: Backlog Behavior (Sales Aggressiveness) Figure 19: Performance (Sales Aggressiveness) Table 2: Performance Measures VariableMeasures CurrentProposed Total Orders164. 07178. 33 Production Backlog2. 91 to 23. 923. 577 to 23. 92 Turn Time1. 164 to 7. 9741. 26 to 7. 974 Outsource Postpress exceeding Capacity Another alternative modeled is to outsource completely any Postpress in-process exceeding capacity. Generally, outsourcing strategy allows organizations to develop and leverage the capabilities required towards global competitiveness (Mclvor, 2008). This significantly decreased the level of production backlog as postpress level is not accumulated overtime having to accomplish finishing requirements outright every time period. Hence, the proposal smoothen overall production backlog without incurring fixed additional costs.
Improved performance is also observed with higher total orders accommodated and less varying backlog and turn time measures. It is also worth noting that increased overall backlog is observed with this proposal. Nonetheless, outsourcing production through subcontracting bindery requirements may be more costly of which effect was not considered in the analysis of the study. Figure 20: Backlog Behavior (Outsource) Figure 21: Performance (Outsource) Table 3: Performance Measures VariableMeasures CurrentProposed Total Orders164. 07225. 19 Production Backlog2. 91 to 23. 927. 141 to 25. 44
Turn Time1. 164 to 7. 9742. 162 to 6. 426 ? Increasing Postprocess Capacity Increasing postprocess capacity against outsourcing option was also considered in case organization would rather fulfill order in-house. Results indicate similar behavioral pattern but with more erratic turn time. Also, overall backlog and turn-time performance are reduced relative to outsourcing option. Figure 22: Backlog Behavior (Postprocess Capacity) Figure 23: Performance (Postprocess Capacity) Table 4: Performance Measures VariableMeasures CurrentProposed Total Orders164. 07223. 37 Production Backlog2. 91 to 23. 27. 715 to 24. 6 Turn Time1. 164 to 7. 9742. 137 to 6. 15 ? 5. Conclusions and Recommendations Production planning has grown increasing interest from logistic managers across companies of different industries. Fundamentally, production process timing and control require close coordination with logistics (Coyle, Bardi, & Langley, 2003). In line with this, computer simulation model using system dynamics was successfully created for a generic print shop. Modeling the causation within production system parameters provided a better understanding of the behavior of these integrated variables.
Alternative scenarios are then simulated to find the path towards higher productivity. The simulation model developed is quite simple with limited variables and feedback structures accounted. Nevertheless, it is comprehensive of a basic print production system presenting the work flow of main offset printing process (prepress, press, and postpress). It is modeled in such a way that it is flexible enough to adapt to any print company putting emphasis on how level of production back¬log affects the turn time of the orders processed and, at the same time, how customer satisfaction reinforces the system.
Although stock flow diagram accounted for multiple variables such as price and quality, it is observed that behavioral pattern of the system is highly influenced by customer satisfaction in terms of turn time or orders, hence the focus on designing alternatives. Sensitivity analysis on capacity changes is provided indicating higher capacity as the answer to turn time minimization as expected. Design alternatives are then centered towards managing demand and considering increase in capacity. Further studies to validate the model and designing alternatives may be considered to continue this research.
The cost function may also be added in the model to assess feasibility of proposals to be generated. ? 6. Bibliography Coyle, J. , Bardi, E. , & Langley, J. (2003). The Management of Business Logistics:A Supply Chain Perspective (7th ed. ). Canada: South-Western: Thomson Learning. Dana, M. (n. d. ). Work with commercial printers to get it right. Retrieved March 03, 2009, from Boston Print Buyers: http://www. bostonprintbuyers. com/articles/article016. html Glykas, M. (2004). Workflow and process management in printing and publishing firms. International Journal of Information Management , 24, 523-538. Grant, A. 2003). Print productivity gap widens. Printing World , 284 (2). Helbing, K. , & Reichel, M. (1998). Selected aspects of development and planning of production and logistic systems. Journal of Materials Processing Technology , 76, 233-237. Introduction to the Offset Printing Process. (n. d. ). Retrieved April 01, 2009, from PsPrint: http://www. psprint. com/resources/printing-tips-and-techniques/general/introduction-to-the-offset-printing-process. asp Kipphan, H. (2001). Handbook of Print Media: Technologies and Production Methods. Springer. Mclvor, R. (2008). What is the right outsourcing strategy for your process?
European Management Journal , 26, 24–34. Merit, D. (1992). Excellence in scheduling print production. New York: Don Merit. Mine, M. (n. d. ). How Offset Printing Works. Retrieved March 28, 2009, from HowStuffWorks: : http://computer. howstuffworks. com/offset-printing. htm/printable Pellow, B. S. (2003). The advertising agency’s role in marketing communications demand creation. Printing Industry Center at Rochester Institute of Technology. Pritschow, G. , & Wiendahl, H. (1995). Application of Control Theory for Production Logistics – Results of a Joint Project. CIRP Annals – Manufacturing Technology , 44 (1), 421-424.
Richmond, B. (2001). An introduction to system dynamics. High Performance Systems, NH. Romano, F. , Fawcett, R. , & Soom, M. (2003). An investigation into printing industry demographics. Printing Industry Center at Rochester Institute of Technology. Uribe, J. (2008). Print Productivity: A System Dynamics Approach. Printing Industry Center at Rochester Institute of Technology. Zeng, J. , Lin, I. , Hoarau, E. , & Dispoto, G. (2009). Numerical Simulation and Analysis of Commercial Print Production Systems. 25th International Conference on Digital Printing Technologies (NIP25).