Journal Search Engine
Search Advanced Search Adode Reader(link)
Download PDF Export Citaion korean bibliography PMC previewer
ISSN : 1598-7248 (Print)
ISSN : 2234-6473 (Online)
Industrial Engineering & Management Systems Vol.21 No.1 pp.74-84
DOI : https://doi.org/10.7232/iems.2022.21.1.074

Improving Daily Total Complete Shipment by Using System Dynamics Simulation in an Indonesian Car Spare Parts Manufacturer

Steven Liang*, Aditya Tirta Pratama, Sumarsono Sudarto
Department of Mechanical Engineering, Engineering Management Concentration, Swiss German University, Indonesia
Department of Industrial Engineering, Swiss German University, Indonesia
Department of Industrial Engineering, Mercu Buana University, Indonesia
*Corresponding Author, E-mail: aditya.pratama@sgu.ac.id
March 2, 2021 July 12, 2021 January 24, 2022

ABSTRACT


PT. Nusahadi Citraharmonis is a leading manufacturing company in Indonesia that produces spare parts for various automotive industries. The company is an active supplier for several well-known car manufacturer in Indonesia, which are: NTC, HMMI, TMMIN, and ASKA. However, the company is currently having difficulty in fulfilling all customer demand on time. Thus, creating backlogged order in the system and decreasing the daily total complete shipment. The objective of this research is to redesign new inventory level to increase the daily total complete shipment. To solve the problem, Define-Measure-Analyze-Improve-Control (DMAIC) is used as the procedure, cause effect analysis is used as a tool to identify the key factors affecting daily total complete shipment, and system dynamics simulation is used as a tool to redesign inventory level. From the result of system dynamics simulation, the newly redesigned inventory level is able to decrease the backlogged order in the system and increase the daily total complete shipment. Besides the system dynamics simulation, standard operating procedure for operating the system is also made and given to the company for future use.



초록


    1. INTRODUCTION

    As a manufacturing company, being able to produce the product on time to fulfil customer’s demand, or usually called just-in-time manufacturing (Pinto et al., 2018), is one the most important aspects to achieve the highest customer satisfaction. To have a smooth production, Sulistyarini et al. (2018) said that a good production planning should be prepared by the company.

    A company should also be able to adapt to several changes in the system that was caused by disruptions (Paul et al., 2016). Howick et al. (2020) mentioned that disruptions might happen in several forms, such as: distribution disruption, machine failure disruption, manpower disruption, supply disruption, and demand disruption.

    PT. Nusahadi Citraharmonis is a leading manufacturing company in Indonesia that produces spare parts for various automotive industries. The company, which was established in 1994 in Pluit, is currently an active supplier for PT. Nusa Toyotetsu Corporation (NTC, sub-assembly manufacturer for Toyota and Daihatsu), PT. Hino Motor Manufacturing Indonesia (HMMI, manufacturer of Hino), PT. Toyota Motor Manufacturing Indonesia (TMMIN, manufacturer of Toyota), and PT. Auto Aska Indonesia (ASKA, sub-assembly manufacturer for Mitsubishi). Located in Cikarang, the main activities of the company are sheet metal shearing, spare parts manufacturing from metal using stamping machine, and spare parts assembling using welding machine and spot machine.

    In January 2020, the company had created 188 times of backlogged order in the system. This number of occurrences of backlogged order, or usually called backlogged order frequency, decreases the daily total complete shipment of the company. By the end of the month, there were 18,186 parts that are failed to be delivered on time. Since then, the company had successfully decreased the monthly backlogged order frequency to only three backlogged order frequency in May 2020. The company found out that the decrease of monthly backlogged order frequency was caused by the decrease of customer demand from January 2020 until May 2020 due to COVID19 pandemic. In addition to that, due to COVID19 pandemic, the company also needed to cut the number of operators (from 500 to 300 operators) in May 2020.

    However, starting from June 2020 until October 2020, the monthly backlogged order frequency had started to increase gradually. The company found out that the increase of monthly backlogged order frequency occurred because the increase of company’s productivity was not as big as the increase of customer demand. Thus, from June 2020 until October 2020, the company had created 231 backlogged order frequency with a total of 27,085 parts that are failed to be delivered on time.

    The company’s main problem is having difficulty in fulfilling all customer demand on time. As a result, there are backlogged order created in the system, which then leads to a decrease in the daily total complete shipment. In other words, inability to fulfill all customer demand on time creates a domino effect to the whole system. The aim of this research is to design new finished goods inventory level to increase the daily total complete shipment of the company.

    2. LITERATURE REVIEW

    This section explains about the frameworks used in this research. Frameworks that could be used to fix the problem are gathered from various references.

    Figure 1 shows the mind map of frameworks used in this research. The literature review mind map consists of: Toyota production system, inventory level control, DMAIC procedure, cause-effect diagram, and system dynamics simulation.

    Toyota production system (TPS) is a type of manufacturing system that was introduced and developed by Taiichi Ohno in Japan (Chiarini et al., 2018). Besides making manufacturing process become more efficient, Mácsay and Bányai (2017) said that by implementing TPS, it will also help company to optimize their manufacturing process. TPS is used as a framework since the company is also implementing TPS in their manufacturing process.

    Inventory level control is a methodology to control the inventory level for just-in-time manufacturing to prevent backlogged order in the system and minimize inventory holding cost (Radasanu, 2016). There are two aspects found in this framework, which are: inventory control calculation (formula to determine the most suitable inventory level) and demand forecasting (formula to foresee the upcoming customer demand). Inventory level control is used as a framework since the company has a difficulty in determining the optimum inventory level.

    DMAIC procedure is a problem-solving method that drives lean six sigma, which focuses on solving the identified problems (Srinivasan et al., 2016). DMAIC procedure consists of five phases, which are: define, measure, analyze, improve, and control (Girmanová et al., 2017;Smętkowska and Mrugalska, 2018). Ongkowijoyo et al. (2020) mentioned that DMAIC is used in their research to identify and reduce lean wastes in a production line of an injection molding company by implementing pull production system. In this research, DMAIC procedure is used as the main approach of this research because the method provides step-by-step process to create improvement in the system, starting from identifying the problem until creating standard operating procedure to maintain the improvement in the future.

    Cause-effect diagram is a diagram used to identify the root causes of the problem (Kuendee, 2017). Besides root cause, cause-effect diagram also mentions the subcauses of each root cause on which aspect does the problem occur. Abbasi et al. (2020) mentioned that causeeffect diagram, or usually called Ishikawa diagram, is used in their research to find the causes of delays in construction industry. Putra et al. (2021) also mentioned that Ishikawa diagram is used in their research to find the causes of having low material efficiency in a tire manufacturing industry. Meanwhile, Pratama et al. (2020) mentioned that Ishikawa diagram is used in their research to find the causes of visual defect in a cable manufacturing industry. In this research, cause-effect diagram is used as a tool to implement DMAIC procedure.

    System dynamics simulation is a tool that is used to analyze the feedback relationship between variables in a closed loop system (Wang et al., 2018). There are four aspects found in this framework, which are: causal loop diagram (diagram to show feedback relationship between variables in the system), stock-flow diagram (diagram to generate dynamic behavior of a system), verification and validation (step to verify and validate the system), and simulation result and sensitivity analysis (output result and disruption testing). The reason of choosing system dynamics instead of basic simulation as a tool for improvement is because system dynamics provides continuous dynamic simulation where the output can affect the input since it has a closed-loop system, while basic simulation only provides discrete event simulation where the output cannot affect the input it has an open-loop system (Raczynski, 2020). System dynamics simulation is also used as a tool to implement DMAIC procedure.

    Besides identifying the frameworks that could be used to fix the problem, most related research about the topic was also gathered. Sánchez-Ramírez et al. (2019) evaluate the impact of machine failure disruptions on the order shipping. The disruption makes the company unable to complete 98-100% order shipping rate. However, by using system dynamics simulation and having on-site spare-parts inventory, the authors were able to find solution to meet the 98-100% order shipping rate.

    Abdullah et al. (2019) evaluate the impact of raw material supply disruption on the backlogged order. The disruption caused the backlogged order increased for double amount in a certain period of time. However, by using system dynamics simulation and having raw material safety stock, the authors were able to decrease the number of backlogged orders in the company.

    Fortunella et al. (2015) evaluate the impact of low production capacity on the backlogged order. The disruption forces the company to increase their production capacity. However, by using system dynamics simulation, the authors were able to decrease the number of backlogged orders in the company.

    Widodo et al. (2010) evaluate the impact of unstable inventory level on the backlogged order. By using system dynamics simulation and redesigning new inventory level, the authors were able to decrease the number of backlogged orders in the company.

    Franco et al. (2020) have a significant backlogged order in their system. However, by using lean six sigma, the authors were able to reduce backlogged hour by 50% and make the company save $370,000 per year.

    3. RESEARCH METHODOLOGY

    Suitable research methods must be prepared before conducting the research. In this section, detailed step-bystep process to conduct the research will be explained. Figure 2 shows the research methodology flowchart of this section.

    As it is shown in Figure 2, since this research is using DMAIC procedure as the main approach, there will be five phases done. In the define phase, problem occur ring in the system will be identified, literatures will be gathered from books and papers, and suitable research methods will be prepared.

    In the measure phase, data related to the variables in the system will be gathered.

    In the analyze phase, root cause problem will be identified using cause-effect diagram. Once the root cause problem has been identified, causal loop diagram will be constructed. After making the causal loop diagram, mathematical formulation will be created and stock-flow diagram will be constructed.

    In the improve phase, verification and validation process will be done. Once the model has is verified and valid, the stock-flow diagram may undergo simulation and sensitivity analysis will be done to test the model.

    In the control phase, standard operating procedure (SOP) for operating the model will be created.

    4. RESULTS AND DISCUSSION

    4.1 Cause-Effect Analysis

    Cause-effect analysis must be done to identify the root causes of the problem, which is having backlogged order in the system. There are two steps done in the cause-effect analysis, which are: cause-effect diagram analysis and cause-effect matrix analysis. Cause-effect diagram and matrix analysis were done in several focused group discussions with the manufacturing leader of the company. Out of seven root causes of the problem, one is identified as the main root cause, which is having low stock of finished goods inventory. Then, to solve the problem, new finished goods inventory level will be redesigned.

    4.2 Causal Loop Diagram Making

    Once the main root cause has been identified, causal loop diagram will be constructed first to show the feedback relationship between variables in the system. Figure 3 shows the causal loop diagram of this research.

    There are two loops found in the causal loop diagram, one balancing loop and one reinforcing loop. In the balancing loop, there are three variables connected, which are: finished goods inventory, kanban, and production rate. A decrease in finished goods inventory will increase the Kanban, which lead to an increase in production rate also. An increase in production rate will then increase the finished goods inventory and create a balancing loop. Desired finished goods inventory has a positive relationship to Kanban and production capacity has a positive relationship to production rate.

    In the reinforcing loop, there are four variables connected, which are: finished goods inventory, delivery rate, backlogged order, and order rate. A decrease in finished goods inventory will decrease the delivery rate, which lead to an increase in backlogged order and an increase in order rate. An increase in order rate will then further decrease the finished goods inventory and create a reinforcing loop. Customer demand has a positive relationship to order rate. Total complete shipment is affected by delivery rate and order rate. Delivery rate has a positive relationship, while order rate has a negative relationship.

    4.3 Stock-Flow Diagram Making

    Once causal loop diagram has been made, stockflow diagram will be constructed first to show the dynamic behavior of the model. This stock-flow diagram will be used as a system dynamics model to run simulation. However, to construct a stock-flow diagram, parameters need to be identified first. Figure 4 shows the stock-flow diagram of this research and Table 1 shows the stock-flow diagram parameters.

    Since the parameters have been identified, then parameters are categorized into two types, input parameters and output parameters. Input parameters are data that will be changed in order to analyze the effect of a certain variable (e.g. customer demand, initial and desired finished goods inventory, and production capacity). While output parameters are data that will be observed as a result of a change in a certain variable (e.g. real time finished goods inventory, backlogged order frequency and number, and daily total complete shipment).

    The input parameters data are gathered from company’s secondary data. This data then will be used further in this research for verification and validation, simulation of system dynamics model, and sensitivity analysis.

    4.4 Verification and Validation

    Once the stock-flow diagram has been made, it needs to undergo verification and validation process first. For the verification process, at the beginning of stockflow diagram development process, there are errors found in the model, such as: negative backlogged order, negative kanban, and total complete shipment above 100%. Fortunately, after several revisions and discussions, there is no more error found in the model.

    For the validation process, there are two tests done, which are: structural test and behavior pattern test. In the structural test, extreme condition testing is done, where value of customer demand as the input parameter is changed into a certain extreme value and value of finished goods inventory as the output parameter is observed.

    In Figure 5, the test shows an expected result where finished goods inventory stays constant for the whole simulation.

    In the behavior pattern test, behavior patterns are compared between proposed model and most related research.

    In Figure 6, the backlogged order behavior pattern of proposed model and Abdullah et al. (2019) is similar, which is growth-and-decline pattern. Meanwhile, in Figure 7, the total complete shipment behavior pattern of proposed model and Sánchez-Ramírez et al. (2019) is also similar, which is decline-and-growth pattern.

    To sum up, since there is no more error found in the model, the structural test shows an expected result, and the behavior pattern test shows the similar behavior pattern with most related research, it can be said that the model fits the real system (Fortunella et al., 2015).

    4.5 Simulation of System Dynamics Model

    Once the system dynamics model is verified and valid, then simulation can be done. But before beginning the simulation, number of samples must be identified first, since there are 147 product types in the population that create a total of 640 backlogged order frequency. By using Pareto principle of 80/20, product types that contribute to 80% of the backlogged order frequency are chosen. As a result, out of 147 population, 65 product types are chosen as the sample population. After determining the number of sample population, simulation is done for each product type. Figure 8 and Figure 9 shows the simulation result of part number 55178-BZ110.

    From Figure 8, in the current condition, there are backlogged order found in January until March 2020. However, in the proposed condition, there is no more backlogged order found for the whole simulation. This shows that there is a decrease in backlogged order in the system.

    From Figure 9, in the current condition, there are decrease of daily total complete shipment in January until March 2020. However, in the proposed condition, the daily total complete shipment stays constant at 100% for the whole simulation. This shows that there is an increase in the daily total complete shipment.

    4.6 Sensitivity Analysis

    After the simulation, sensitivity analysis is done for several trials. The value of production capacity as the input will be decreased and daily total complete shipment as the output will be observed.

    The orange highlighted area in Table 2 shows the condition when there is no production capacity decrease in the simulation. Then, to test the sensitivity of the system dynamics model, production capacity is decreased by 2.5% for each trial. In total, there are 20 trials for the sensitivity analysis, with a total of 50% production capacity decrease.

    In Table 2, it is also shown that the minimum total complete shipment for trial 1-17 stays constant at 100%. In trial 18-20 however, it starts to decrease, shown by the green highlighted area.

    In trial 18, when the production capacity is decreased by 45% to 440 pcs/day, there are 4 backlogged orders found in the system, with a total of 290 parts. Besides, the minimum total complete shipment also decreases to 88.1%. In trial 19, when the production capacity is decreased by 47.5% to 420 pcs/day, there are 19 backlogged orders found in the system, with a total of 4,485 parts. Besides, the minimum total complete shipment also decreases to 41.2%. In trial 20, when the production capacity is decreased by 50% to 400 pcs/day, there are 46 backlogged orders found in the system, with a total of 19,009 parts. Besides, the minimum total complete shipment also decreases to 0%.

    Thus, from the sensitivity analysis, it shows that the newly designed inventory level for this part number has 42.5% production capacity decrease threshold (trial 17) to maintain 100% minimum daily total complete shipment.

    4.7 System Dynamics Model Manuals

    To maintain the use of the proposed system dynamics model, standard operating procedure for operating the system dynamics model will be made. With this system dynamics model manuals, operators can rely on this document whenever problem occurs in the future. In total, there are two files that will be given to the company, which are: optimum inventory calculator (filetype: Excel) and system dynamics simulation (filetype: PowerSIM). These files will be given and explained during the training process.

    5. CONCLUSION

    This research shows the combination of using causeeffect diagram and system dynamics simulation as tools in implementing DMAIC procedure for company’s improvement. The findings of this research are shown in Figure 10.

    As shown in Figure 10, 46 out of 65 samples increase its minimum daily total complete shipment. Before the newly designed inventory level is implemented, the average minimum daily total complete shipment is 51.3%. But after the newly designed inventory level is implemented, the average minimum daily total complete shipment increases to 74.9%. In addition to that, 55 out of 65 samples decrease its backlogged order frequency, and 50 out of 65 samples decrease its backlogged order number..

    This research contributes to the academic area by establishing new perspective for improving daily total complete shipment in an automotive manufacturing industry. This research creates a new perspective in solving the problem, since this research uses DMAIC as the procedure and system dynamics simulation as a tool to implement DMAIC procedure.

    In addition to that, this research also contributes to the practical area by showing improvement of daily total complete shipment through the use of system dynamics simulation.

    6. RECOMMENDATION

    For future research, since the development of system dynamics simulation in this research focuses only on redesigning the inventory level, it is better to also consider the production planning since the company is producing various types of spare parts and there is a production capacity for producing those spare parts in a day. By also taking production planning into consideration, the output produced will become more valid.

    ACKNOWLEDGEMENT

    I would like to thank Mr. David as the owner of PT. Nusahadi Citraharmonis for giving me the opportunity to do this thesis research in his company and sharing the data needed for this thesis research.

    I would like to also thank Mr. Sumarno as the manufacturing leader of the company and my supervisor during this thesis research.

    Figure

    IEMS-21-1-74_F1.gif

    Literature review mind map.

    IEMS-21-1-74_F2.gif

    Research methodology flowchart.

    IEMS-21-1-74_F3.gif

    Causal loop diagram.

    IEMS-21-1-74_F4.gif

    Stock-flow diagram.

    IEMS-21-1-74_F5.gif

    Extreme condition test result.

    IEMS-21-1-74_F6.gif

    Behavior pattern test (backlogged order).

    IEMS-21-1-74_F7.gif

    Behavior pattern test (total complete shipment).

    IEMS-21-1-74_F8.gif

    Simulation result of part number 55178-BZ110 (backlogged order)

    IEMS-21-1-74_F9.gif

    Simulation result of part number 55178-BZ110 (daily total complete shipment).

    IEMS-21-1-74_F10.gif

    Output result summary.

    Table

    System dynamics simulation parameter

    Sensitivity analysis of part number 51266-0K010

    REFERENCES

    1. Abbasi, O. , Noorzai, E. , Gharouni Jafari, K. , and Golabchi, M. (2020), Exploring the causes of delays in construction industry using a cause-and-effect diagram: Case study for Iran, Journal of Architectural Engineering, 26(3), 4-14.
    2. Abdullah, M. A. , Hishamuddin, H. , and Bazin, N. (2019), A system dynamics approach to investigate the effects of disruption on the supply chain with a mitigation strategy, IOP Conference Series: Materials Science and Engineering, 697(1), 012024.
    3. Chiarini, A. , Baccarani, C. , and Mascherpa, V. (2018), Lean production, Toyota production system and Kaizen philosophy: A conceptual analysis from the perspective of Zen Buddhism, The TQM Journal, 30(4), 425-438.
    4. Fortunella, A. , Tama, I. P. , and Eunike, A. (2015), Model simulasi sistem produksi dengan sistem dinamik guna membantu perencanaan kapasitas produksi, Jurnal Rekayasa dan Manajemen Sistem Industri, 3(2), 256-267.
    5. Franco, D. , Henry, J. , Mulligan, A. , Schiher, B. , Pegues, K. , and Enos, J. (2020), Lean six sigma: Reduction of backlog hours within the equipage division, 2020 Annual General Donald R. Keith Memorial Capstone Conference, West Point, New York, USA, 101-107.
    6. Girmanová, L. , Šolc, M. , Kliment, J. , Divoková, A. , and Mikloš, V. (2017), Application of six sigma using DMAIC procedure in the process of product quality control in metallurgical operation, Acta Technologica Agriculturae, 20(4), 104-109.
    7. Howick, S. , Ackermann, F. , Eden, C. , and Williams, T. (2020), Delay and disruption in complex projects, In: B. Dangerfield (ed.) System Dynamics. Encyclopedia of Complexity and Systems Science Series, Springer, 315-319.
    8. Kuendee, P. (2017), Application of 7 quality control (7 QC) Tools for quality management: A case study of a liquid chemical warehousing, 20174th International Conference on Industrial Engineering and Applications (ICIEA), 106-110.
    9. Mácsay, V. and Bányai, T. (2017), Toyota production system in milkrun based in-plant supply, Journal of Production Engineering, 20(1), 141-146.
    10. Ongkowijoyo, Y. K. , Pratama, A. T. , and Sofianti, T. D. (2020), Supply chain improvement in a local toy manufacturing company by a kanban system, Proceedings of the International Conference on Engineering and Information Technology for Sustainable Industry, 1-6.
    11. Paul, S. K. , Sarker, R. , and Essam, D. (2016), Managing risk and disruption in production-inventory and supply chain systems: A review, Journal of Industrial and Management Optimization, 12(3), 1009-1029.
    12. Pinto, J. L. Q. , Matias, J. C. O. , Pimentel, C. , Azevedo, S. G. , and Govindan, K. (2018), Introduction to Lean and Just-in-time Manufacturing, In Just in Time Factory, Springer, 1-4.
    13. Pratama, A. T. , Saraswati, T. , Prianggara, F. , and Savitri, T. (2020), Improving productivity and quality of medium voltage cable production, IOP Conference Series: Materials Science and Engineering, 847(1), 1-8.
    14. Putra, A. R. , Sudarto, S. , and Pratama, A. T. (2021), Improving material efficiency on calendering process by using six sigma and define measure analyze improve control (DMAIC) methods in Indonesian leading truck bus bias tire manufacturer, Proceedings of The Conference on Management and Engineering in Industry, 3(2), 1-6.
    15. Raczynski, S. (2020), Interacting Complexities of Herds and Social Organizations, Springer, Singapore.
    16. Radasanu, A. C. (2016), Inventory management, service level and safety stock, Journal of Public Administration, Finance and Law, 9,145-153.
    17. Sánchez-Ramírez, C. , Ramos-Hernández, R. , Mendoza Fong, J. R. , Alor-Hernández, G. , and García-Alcaraz, J. L. (2019), A system dynamics model to evaluate the impact of production process disruption on order shipping, Applied Sciences, 10(1), 208-222.
    18. Smętkowska, M. and Mrugalska, B. (2018), Using six sigma DMAIC to improve the quality of the production process: A case study, Procedia-Social and Behavioral Sciences, 238, 590-596.
    19. Srinivasan, K. , Muthu, S. , Devadasan, S. R. , and Sugumaran, C. (2016), Six sigma through DMAIC phases: A literature review, International Journal of Productivity and Quality Management, 17(2), 236-257.
    20. Sulistyarini, D. H. , Femiana, G. , and Patriotik, N. (2018), Pemerataan produksi culvert dengan metode heijunka, Jurnal Ilmiah Teknik Industri, 5(3), 171-178.
    21. Wang, D. , Nie, R. , Long, R. , Shi, R. , and Zhao, Y. (2018), Scenario prediction of china’s coal production capacity based on system dynamics model, Resources, Conservation and Recycling, 129, 432-442.
    22. Widodo, E. M. , Fatimah, Y. A. , and Indarto, S. (2010), Simulasi sistem dinamik untuk meningkatkan kinerja rantai pasok (Studi kasus di industri kulit pt lembah tidar jaya magelang), Industrial and Systems Engineering Assessment Journal (INASEA)-Discontinued, 11(1), 35-44.
    Do not open for a day Close