Abstract
Companies are increasingly operating in a dynamic environment that requires them to adapt to constant fluctuations in demand and improve productivity while continually reducing costs. This paper reports a case study of a manufacturing company that needed to improve its production planning and control of semi-finished products. During COVID-19, there was a significant increase in orders, which put a great deal of pressure on the production system and led to high downtime on the assembly lines and backlogs due to component shortages. Once the problem was characterized and the root causes identified, the communication between the planning and production departments was improved using process modeling with BPMN (Business Process Modeling and Notation), and a dynamic Kanban system was implemented in the most critical area to align production with demand, based on the principles of Lean Manufacturing. The measures implemented helped to immediately reduce the downtime on the assembly lines, the intermediate stocks and the amount invested. This work contributes to the literature by reporting the application of an advanced Lean tool such as Kanban, and by combining two philosophies for process improvement with the joint application of BPMN and Kanban. It may also support other companies facing the same problems by providing guidelines for using and implementing these tools.
Keywords:
Production planning and control; Lean Manufacturing; Kanban; Business Process Modeling and Notation; Information Management
Resumo
As empresas operam cada vez mais em um ambiente dinâmico que exige a capacidade de se ajustarem às constantes flutuações na demanda e à necessidade de melhorarem continuamente sua produtividade, reduzindo custos. Este trabalho relata um estudo de caso de uma empresa de manufatura que precisava melhorar o planejamento e controle da produção dos produtos semi-laborados. Durante a COVID-19, verificou-se um aumento significativo dos pedidos, o que criou grande pressão sobre o sistema produtivo, elevados tempos de espera nas linhas de montagem e trabalho acumulado por falta de componentes. Depois de caracterizado o problema e identificadas as causas-raiz, foram implementadas melhorias na comunicação entre os departamentos de planejamento e produção. Para isso, recorreu-se à modelagem de processos com BPMN (Business Process Modeling and Notation) e foi criado um sistema Kanban dinâmico no setor mais crítico, procurando alinhar a produção com a demanda, tendo por base os princípios da produção Lean. As medidas implementadas permitiram diminuir os tempos de espera das linhas de montagem e reduzir os estoques intermediários e o respectivo valor investido na sua manutenção. Este trabalho contribui para a literatura ao relatar a aplicação de uma ferramenta Lean avançada como é o caso do Kanban e combinar duas filosofias diferentes, mas que se podem complementar para a melhoria de processos, com a aplicação conjunta da BPMN e Kanban. Pode também servir para apoiar outras empresas que enfrentem o mesmo tipo de problemas, fornecendo-lhes diretrizes para a utilização e implementação deste tipo de ferramentas.
Palavras-chave:
Planejamento e controle da produção; Produção Lean; Kanban; Business Process Modeling and Notation; Gestão de Informação
1 Introduction
Companies must continuously seek to improve their productivity while reducing costs in order to remain competitive. Lean Manufacturing (LM) is among the continuous improvement approaches that have gained the most attention in recent decades (Maldonado et al., 2020). This philosophy aims to identify and eliminate non-value-adding activities, reduce waste, and standardize processes (Womack & Jones, 2003). Most studies on implementing LM report a positive impact on operational, financial, and environmental performance (Negrão et al., 2017). However, some authors argue that the lack of a common framework for implementing LM contributes to a high failure rate (Mamoojee-Khatib et al., 2023). The adoption of Lean tools in manufacturing companies often occurs randomly and in an unstructured manner, based on the particularities and problems of each company (Maware et al., 2022). Additionally, according to a literature review conducted by Badhotiya et al. (2024), most published case studies report the application of basic tools such as Value Stream Mapping (VSM) or 5S, which can be used in any production system and are associated with an earlier phase of implementation. On the other hand, the implementation of tools such as SMED (Single-Minute Exchange of Die), Heijunka, or Kanban, associated with more mature projects and make-to-order production systems, has been less reported in the literature. This happens because companies primarily tend to implement simpler practices that generate short-term results, neglecting the use of tools that require greater implementation efforts (Negrão et al., 2017; Badhotiya et al., 2024) and the benefits that can arise from integration with other process improvement philosophies (Mamoojee-Khatib et al., 2023). Business Process Management (BPM) is one such philosophy. This philosophy focuses on improving workflows using tools such as Business Process Modeling and Notation (BPMN), which allows for visualizing communication flows between actors, task sequence, and task assignment (Santhiapillai & Ratnayake, 2024).
Thus, there is an opportunity to conduct research that reports implementations with a different level of maturity, integrating other management philosophies and, consequently, helping to identify implementation patterns that can guide other companies in improving their processes. This work aims to address this need by presenting a case study of a manufacturing company where, initially, BPMN was used to implement improvements in communication between the Planning and Production departments. Subsequently, a Kanban system was created to align production with demand and improve the factory's planning and control of semi-finished products (SFPs). The joint implementation of these two approaches reduced supply failures to the assembly lines and the accumulated downtime due to these failures.
The aim of this research is to contribute to LM literature by presenting a case study with the application of an advanced Lean tool (going beyond the implementation of basic tools), preceded by the mapping of information flows with BPMN, demonstrating the feasibility of using these two tools together for process improvement. This study can also offer important guidelines regarding the process of jointly implementing various tools, particularly BPMN and Kanban, supporting companies facing similar problems that want to implement a pull system to adjust production to demand.
The article is structured as follows. In the next section, a brief theoretical framework is provided. Subsequently, the case study methodology is presented, followed by a description of the work carried out on the shop floor. Finally, the results are discussed, and the article concludes with the contributions of the work, limitations, and opportunities for future research.
2 Theoretical background
2.1 Lean Manufacturing
LM is a management philosophy that seeks to identify and eliminate non-value-adding activities, reduce waste, and standardize processes (Womack & Jones, 2003). This philosophy focuses on the physical flow of materials and is based on the collaboration and commitment of everyone to create and maintain a culture of continuous improvement (Maldonado et al., 2020). LM has become popular in manufacturing companies (Badhotiya et al., 2024; Maware et al., 2022) with proven results in reducing cycle time, waste, and inefficiencies, as well as improving product quality, delivery time, worker productivity, and lead time (Maware et al., 2022).
The implementation of LM is based on five fundamental principles established by Womack & Jones (2003): (1) define the value of a product from the customer's perspective; (2) identify the value stream and eliminate waste; (3) create continuous flow wherever possible; (4) transition from a push production system to a pull system, producing and delivering what the customer wants only when they want it; and (5) seek perfection through continuous improvement. This implementation can be supported by tools and techniques designed to reduce different types of waste and meet different objectives (Badhotiya et al., 2024; Naeemah & Wong, 2023). According to Tortorella et al. (2015), more than 100 tools and techniques exist. This number is constantly growing, the selection of which being motivated mainly by usefulness in each situation and potential impact.
Badhotiya et al. (2024) assess the maturity of LM implementation based on the tools used by the company and the extent to which they contribute to implementing the five fundamental principles. They concluded that more generic tools (e.g., VSM, 5S, or work standardization), which are used to eliminate waste and applied regardless of the type of production system, are associated with less mature implementations and more focused on the first three principles. In contrast, the use of tools such as Kanban or SMED is associated with more mature implementations aimed at achieving the fourth principle. Companies tend to prioritize the rapid implementation of simpler practices that lead to more tangible short-term results (Negrão et al., 2017), favouring the use of more generic tools and facing more difficulties in implementing more advanced tools, which leads to a stabilization of the performance improvement rate (Badhotiya et al., 2024).
2.2 Kanban system
The Kanban system is considered a tool designed for managing production flows. It supports the creation of a pull system, advocated by the fourth fundamental principle of LM. It allows flexible and organized communication between factory sectors, forcing the downstream sector to indicate to the upstream sector what needs to be produced and in what quantities (Ohno, 1988). It is used as a production planning and control tool in assembly production systems, where the demand for components depends on the market demand for the final product (Thürer et al., 2022). Thus, the Kanban system tends to be implemented in organizations that make-to-order and seek flexibility as part of their organizational competencies (Badhotiya et al., 2024).
This type of system was initially developed for deterministic production environments with stable demand and processing time. However, demand uncertainty is an increasing reality in the current context, and there may be a need to adjust the number of Kanban cards in the production system according to demand variation. This has led to the development of dynamic or flexible Kanban systems instead of fixed-card Kanban systems used in stable-demand environments (Guneri et al., 2009; Reddy et al., 2023).
Kanban, in its most widely used version, is a system that uses so-called supermarkets, or controlled inventory locations, that are supplied based on cards. These cards signal which components have been consumed and the quantities that need replenishment (Thürer et al., 2022). This contributes to controlling intermediate stocks, avoiding overproduction and preventing stocks from falling below minimum levels (Rahman et al., 2013). Other advantages include simplicity in production planning, reduced operator workload, ease of identifying components by the Kanban card attached to containers, and substantial reduction in bureaucracy (Guneri et al., 2009).
Gross and McInnis (2003) suggest seven steps for correctly implementing a Kanban system. The first step involves collecting the necessary data to characterize the process and make fact-based decisions. The Kanban's size is calculated in the second step, considering the current system conditions. The number of Kanban cards associated with a particular component will depend on demand, lead time, safety stock, and container capacity (Reddy et al., 2023). Subsequently, the rules for information and material flows are established, resulting in an implementation plan. In the fourth step, the project and its rules are presented to the team, who conduct some simulations if possible. They then verify whether all conditions are met for the start of Kanban. In the last two phases of the process, auditing and maintenance are carried out, controlling the system's progress and verifying if it meets its needs. Possible improvements to the system are identified and analyzed.
2.3 Business Process Management
Similarly to LM, BPM is a management philosophy used in process improvement. Dumas et al. (2018) define it as the art of understanding how work is done in an organization to ensure consistent results and take advantage of improvement opportunities. In a more traditional view, a company is organized by functional areas or departments, where each develops plans to implement specific improvements. However, these may overlook the needs of their peers (Ferreira et al., 2018). The BPM philosophy allows organizations to have a broader view of their processes, considering all those involved (Dani et al., 2019), which helps identify points where information or work is transferred (Ferreira et al., 2018), deconstruct complex processes and increase knowledge about them (Recker, 2010).
Several authors identify the need for organizations to redesign their processes before implementing LM (Maware et al., 2022; Negrão et al., 2017). The BPM philosophy and BPMN, as its reference notation, can complement a Lean approach by providing a more detailed view of information flows and the various stakeholders in the processes, serving as a basis for eliminating waste associated with information management (Calçado et al., 2024). Additionally, BPMN is recognized as a graphical notation easily understandable by all users, from analysts to those responsible for process implementation (Chinosi & Trombetta, 2012), unlike VSM, which, although more frequently used in LM implementation, requires greater knowledge of the notation and is used primarily to map material flows (Soliman et al., 2022).
However, despite the applicability of these tools to process improvement, examples of combining these management philosophies with others are still scarce (Calçado et al., 2024; Ferreira et al., 2018), especially in manufacturing companies, where little is known about the joint application of BPMN with more advanced Lean tools such as Kanban.
3 Methodology
This article reports a case study of a company that applied BPMN and Kanban to improve the production planning and control of SFPs, which are then integrated into the final products. Case studies are suitable when the focus is on a contemporary phenomenon in a real context (Yin, 1994). They are helpful because they offer a comprehensive understanding of the nature and complexity of the phenomenon, which is acquired through observation, having great validity for professionals who are the end-users of the research (Voss et al., 2002). Case studies are widely used in operations management (e.g. Calçado et al., 2024) as a practical technique to assess the applicability of tools and methods to improve company performance (Kitchenham et al., 1995).
The factory under study belongs to a German group with 15 factories spread across Europe, Asia, and America, producing more than 15 million products annually. The factory is primarily dedicated to the production of household appliances. The need for intervention in production planning and control arose mainly from the increase in sales during the COVID-19 pandemic, which put pressure on all sectors, especially production, which supplies the assembly lines, leading to a high number and duration of stoppages. Thus, the first step was to better characterize the problem by collecting data and identifying root causes, followed by the implementation of some improvements in the sector considered most critical to address the issues identified. These improvements consisted of enhancing information flows between teams supported by BPMN and aligning demand and production through implementing a Kanban system according to the methodology proposed by Gross & McInnis (2003).
4 Case study
4.1 Problem description
In the second half of 2020, the factory under study faced a sharp increase in the number of orders, causing great pressure in all areas, from the Sales department to Production, exceeding the production capacity for which it was prepared. The factory produces many components integrated into the final products, known as SFPs. For this, there are several manufacturing sectors in the factory: lasers, bending machines, panel bending machines (Computer Numerical Control (CNC) bending machines), mechanical and hydraulic presses, welding, painting, screen printing, gluing, and reprographics. An SFP may pass through only one of these manufacturing sectors. However, it may also pass through several, and not all follow the same sequence, making it difficult to have a product-based layout and requiring effective communication and organization to ensure nothing fails between these manufacturing sectors. The increase in sales led to a backlog in orders, partly due to a lack of installed capacity, a high number of assembly line stoppages and high downtime. This became a problem in terms of production planning and control.
Information was collected from two sources to identify the reasons behind assembly line stoppages. The information written by the operators on the OEE (Overall Equipment Effectiveness) sheets and the information gathered through daily conversations with the assembly line supervisors. Additionally, the number of assembly line stoppages and the accumulated downtime were analyzed. This analysis found that in the worst week of October 2020, downtime due to lack of SFPs was close to 3,500 minutes or about 58 hours.
Next, the information collected as regards the root causes was represented in an Ishikawa diagram. The Ishikawa diagram (or cause-and-effect diagram) highlights the relationship between qualitative attributes and their related factors. The problem is identified on the main line, and all the causes found for the problem are presented on the branches that connect to the main line, sometimes having other associated causes. With the information presented in this way, team members can address the various leading causes of the problem (Yazdani & Tavakkoli-Moghaddam, 2012). Figure 1 shows the Ishikawa diagram. The identified causes were grouped by the three main manufacturing sectors – LPB (Lasers, Panel Benders, Bending Machines), PSR (Painting, Screen Printing, and Repairs), and presses – by the three areas of the Supply Chain department – Purchasing, Planning, and Warehouse of raw materials (RM) and consumables (CM) – and by the assembly lines and supply.
As can be observed, in all manufacturing sectors, the lack of capacity was identified as one of the main reasons why they could not deliver the SFPs on time for assembly. The lack of production control was also noted in the LPB manufacturing sector, where "not following the plan" and "overproduction" were identified as causes of the problem. The LPB sector was chosen for intervention because it is one of the first stages in SFP manufacturing (it receives the metal sheet directly from the supplier and cuts it into different shapes). In the case of delays, it contributes to increasing the risk of production delays in the rest of the factory.
4.2 Implementation of planning improvements in the LPB sector
4.2.1 Communication between Planning and Production teams
Before implementing a Kanban system in the LPB sector to adjust the manufacturing of SFPs to the demand in the assembly lines, the planning process in this sector was analyzed in order to identify inconsistencies that might be contributing to the identified problems. For this, BPMN was used to model the process, considering the various stakeholders and the steps to obtain the weekly sequencer of the SFPs required on the assembly lines and their respective quantities (Figure 2).
As illustrated, there are two planners, one responsible for the exhaust and chimney area and one responsible for the cavity area, who each prepare a general monthly production plan after receiving the sales orders for the following month. Weekly, each planner settles a weekly plan for their own area. Based on these two weekly plans, an availability analysis is carried out in the company's system. This analysis results in a list of all the SFPs needed to meet demand, which is then updated to an Excel file used to create the production sequencer. However, the planner for the exhaust area also updates an Excel file with the number of units needed per day for the next two weeks. This information only refers to some of the highest turnover SFPs and not all of them.
After analyzing the process, a potential reason for the non-compliance with the plan (sequencer) was identified, which resulted in overproduction. The reason was that the information provided by the weekly exhaust plan was used in two different sequencers. After talking to the operators who used the sequencers in their work, it was concluded that this was detrimental and created confusion and information conflicts because they did not know which plan to follow. This resulted in overproduction and non-compliance with the plans (represented in the Ishikawa diagram). Therefore, the information about the exhaust area was removed from the common sequencer to have a separate file for each manufacturing area instead.
Additionally, the availability analysis took too long, often resulting in the sequencer being delivered the day after the planning period began. This led to outdated information and at least one day of unplanned production. Since this file was only used for the cavities area, a support file was created containing the Bill of Materials (BOM) for all finished products, allowing the weekly plan to be loaded and immediate access to a list of SFPs that would be needed, without having to wait for the results of the availability analysis. This file is used, but always with the commitment of the person responsible to regularly update the BOMs and make the necessary changes when there is any change in a product. After these changes, BPMN was used again, and the TO-BE model of the process was developed (Figure 3).
During the process analysis, another problem was identified: the sequencer is static and quickly becomes obsolete with the various changes that can occur in the assembly line plans. The file cannot analyze the stock of SFPs, so it analyzes each manufacturing sector individually. Thus, what can happen is that, for example, an SFP is requested on laser cutting machines when the stock, already bent, is sufficient for the required production.
This analysis must be carried out by the team leaders responsible for the LPB sector to confirm that they will not use raw materials and time that could be spent on other activities. Since this analysis must be done by employees who do not have access to the company's system to see the codes of the following sectors, a "Next Code" column was added to the Laser machines sheet and a "Previous Code" column to the others (bending machines and panel benders). Additionally, an informal meeting was held with the team leaders of both shifts and their supervisor to understand how much the sequencer helped them and what fields could be added or removed. As a result of this meeting, the "Day," "Work Center," and "Production Order" columns were removed from the document, and the "Observations" column was replaced with one to track the quantity already produced that week.
4.2.2 Definition of minimum quantities for manufacturing orders
During the analysis of the planning process using BPMN, other issues were identified, such as overproduction, which generated a lack of raw materials and impacted the problem under study. Often, instead of following the quantities on the weekly sequencers delivered, employees produced quantities that appeared random, especially for small-sized parts. However, after questioning them, it was found that, for example, if a manufacturing order for the laser cutting machines was for 500 units, 3100 units would be manufactured, as it is the number of parts that fit on a whole metal sheet. This is because if the cutting were adapted to the required quantity, there would be scraps of different sizes for which there are no defined cutting programs. This would mean the Design department must create new programs to use these scraps whenever necessary.
This situation directly affects purchasing technicians, as they purchase raw materials to produce only the required units, leading to a shortage of raw materials for SFPs that use the same metal sheet specification. To prevent these cases, minimum production order quantities were introduced, which makes the system recommend the purchase of metal sheets considering these values. To obtain the number of parts per sheet, the Design department was contacted, and operators were asked to note the number of parts manufactured during standby time (while the machine was cutting). These values were subsequently loaded into the system, helping to define minimum production order quantities. This not only helps the purchasing technicians prevent the lack of raw materials as the purchase considers the minimum order quantities but also allows the production supervisor to open production orders with realistic quantities, which helps to know the real time required to manufacture a part.
4.2.3 Implementation of a Kanban system in the sector
After using BPMN to model and identify opportunities for improvement in the production planning process in the LPB sector, a pilot project for the Kanban system was implemented to intervene at the production control level and address some of the problems identified. For this, the seven steps for implementing a Kanban system by Gross & McInnis (2003) were followed.
4.2.3.1 Data collection and Kanban size calculation
To begin, more specific data about the manufacturing sector was collected. The 2020 consumption data was used to perform an ABC analysis of the SFPs passing through the manufacturing sector. Considering production volumes, this analysis allowed all SKUs (Stock Keeping Units) to be organized into three general categories (A, B, and C). SKUs representing up to 70% of manufactured parts were considered in Category A, up to 90% in Category B, and the remainder in Category C.
This analysis allowed categorizing 26 of the 292 SFPs that passed through the bending machines as Category A, 33 of the 249 through the laser and 10 of the 62 through the panel bending machines. Considering this set of SKUs, a table was created to calculate the safety stock for each of them. Initially, this list was analyzed to ensure that the SKUs formed groups since many items were included in the same finished product but did not have the same consumption values.
In addition to this analysis, several Gemba walks helped to understand the flow of the SKUs categorized as A through the factory and the capacity of the containers. A list was made because there was no established standard or database with this data, which was necessary to calculate the Kanban size.
The safety stock was calculated for each SKU by multiplying the risk of stockout by the demand variance during the lead time to calculate the Kanban size. The continuous review model was considered. According to this model, when the stock of an item reaches a predefined minimum value, it is necessary to produce more to return to the maximum stock level. If this does not happen, the risk of stockout increases. After obtaining the values for the safety stock, the other data necessary to determine the number of Kanban cards for each SKU (N) were obtained, namely the average demand per unit of time (µr), the average replenishment time (L) for delivery of components, and the capacity of the containers used to store the SKUs (C). Equation 1 was used, as Chan (2001) suggested.
Consumption values from the second half of 2020 were considered to calculate the average demand, and the average replenishment time was one full day (two production shifts).
4.2.3.2 Kanban design
At this stage, all the operational details of the Kanban system have been defined, such as what visuals could be implemented or who decides what to produce when there are scheduling or capacity conflicts.
The Kanban system was established in the bending machine area with only four SKUs. For some aspects of the Kanban design, such as defining inventory locations and the flow of information between suppliers and team leaders, meetings were held with everyone who used it. Including these people on the team contributed to their increased commitment to the project, resulting in more attention paid to every detail, greater dedication and effort undertaken to ensure nothing went wrong. Moreover, it prevented the choosing of locations that did not make sense and allowed adjustments in quantities based on their knowledge. It was decided that four pallets with fixed locations would be placed next to the bending machines to test the system. When a supply operator comes to pick up a container with material, the card is left in the right place as well as is an empty container (Figure 4). By leaving this empty container, they can visually manage existing inventory and avoid using cardboard boxes to store SFPs, which pose safety risks. In addition to these rules, it was decided that any production planning conflicts would be reported to the production supervisor and, if necessary, to the production planning technicians. It was also agreed that the supervisor would audit the Kanban system once a week, checking with operators if there were any issues, with the freedom to increase the time between audits later.
In laser and panel bending machines, the design of the Kanban system was slightly different, as these two areas are in the same manufacturing sector. Only the team leaders of the three shifts were invited to participate. Their participation provided important inputs on the available space for the Kanban stock storage, which was considered in order to limit the number of SKUs to be included in the Kanban system (until more space could be freed up) and to critically evaluate the calculated batch quantities (with the aim of reducing them as much as possible). Due to the lack of space, it was not possible to replicate the visual management method implemented in the bending machine area (i.e., using empty containers indicating the need for manufacturing), so the system shown in Figure 5 was created.
Initial visual management board for the Kanban system of the laser and panel bending machines
Magnets must be placed on the board when a Kanban card is left on a batch of manufactured material and removed when the next sector picks up that batch, leaving the card in the green container next to the board. Red squares alert workers to the fact that an SKU is reaching the safety stock, and a red line identifies the maximum stock. Thus, all inventory information is in one place. Additionally, next to it is the shelf map so as to allow the checking of quantities when auditing the system.
Although the design of the Kanban system differed, the rules regarding conflicts and audits remained the same as in the bending machine area. However, team leaders have more freedom as most information flows within the sector. One of the most important rules imposed was regarding the return of the Kanban card. Once the SFPs leave the LPB sector after being processed on bending machines and panel benders, supply operators immediately leave the card in the correct location when picking up the SKUs (in pallet or container), which means that the stock is considered consumed for the sector from that moment on. However, for cards relating to material cut on laser machines, as they remain within the sector itself (mostly going to panel bending machines), the card will only be returned when all the stock for that batch has been consumed.
4.2.3.3 Training and system implementation
In the case of the bending machine area, two Kanban initiation sessions were held: one for the team leaders of each of the three shifts and their operators and another for the supply teams (only those people who had worked with the SFPs covered by the Kanban system), always providing all the help and time necessary for the concept to be well understood. After that, a week of intensive monitoring of all participants and SKUs began to ensure that the SFPs were not missing on the assembly lines during the adaptation phase and that everyone involved understood and followed the rules.
In laser machines, the same procedure was followed. All operators in this area were trained. It was explained to them how the system would work, its role, and its importance. At this stage, the panel bending machine operators were also trained, and training for the bending machine operators was reinforced, as they were the ones who used the material cut by the laser machines. These efforts in training the bending machine operators were necessary because they play different roles in the system. Accordingly, they produce and place the cards in the bending machine area and use the material from laser machines, having to deliver the cards to trigger the production order.
4.2.3.4 Audit, maintenance, and improvement of the Kanban system
After three weeks of testing the system with a few SKUs, operators understood how it worked and how to work with it (with only a few occasional mistakes). Therefore, it was decided to expand the system to include more SKUs and panel benders. All rules remained the same; however, it was necessary to increase the space for the visual management system (Figure 6) and train the supply operators responsible for collecting and gathering the material manufactured in the panel benders. All support was provided to ensure the system worked correctly and that communication flowed between the various sectors. The assembly and paint line operators were trained as they consumed the SKUs added to the Kanban system.
After its implementation, the Kanban board underwent small changes to improve the flow of information, with the most impactful change being the development of One Point Lessons (OPL), both to support operators and to train new operators in the future. These OPLs included updating the board by those who produced and consumed the SFPs. An audit support sheet was also created for the Kanban system, allowing a quick and general analysis of its functioning and identifying anomalies, such as lost cards, before they caused significant problems on the assembly lines.
Over time, with the implementation of the Kanban, there were some struggles in controlling minimum and maximum stock levels in the panel benders. It was found that the issue was that some SKUs, when leaving the panel benders, could not go directly to the Kanban as they still needed to be riveted and could only go to assembly lines after this process. This procedure had to be performed on six SKUs in the system. However, this constituted an issue for only three of them because the production rate of these SFPs on the machine was higher, and the workers could not operate the machine and rivet simultaneously without allowing the accumulation of parts at the machine's exit. Since there was a need to first store pallets of parts for riveting and then, when available, to rivet them, the team used this as an excuse to overproduce since, technically, the Kanban system did not cover them. This problem arose due to the lack of specific rules regarding this procedure. This ended up being an opportunity to improve the Kanban system and to create new rules, improve the information and material flows, and increase team commitment by involving it in the development of new rules.
The visual management board for the Kanban system with an additional section for pallets to be riveted is shown in Figure 6. A magnet is placed even when batches of material to be riveted are stored, and a Kanban card is not placed because they are not ready to go to the assembly line. Therefore, the production of these SKUs is limited, as the total number of batches manufactured (those to be riveted and those already riveted) must be equal to the total number of magnets on the board, which correspond to the quantities established in the Kanban system.
During this process, the operators took the opportunity to show some concerns regarding the production sequencing procedure for the SKUs covered by the Kanban system. The FIFO (First In, First Out) rule was applied, except when an SFP had increased demand and went into red alert (safety stock), making it necessary to manufacture it first. However, by following this rule, stocks sometimes became critically low. This happened because the consumption rate of some SFPs is higher than that of others, and the demand is not levelled. A daily plan with information on which SFPs should be consumed on the day started to be provided to assist the team leaders in sequencing production in case of doubts.
In addition to the adjustments made to the physically implemented Kanban system, there was an additional need to create a tool to assist in its maintenance and make it dynamic, allowing the reassessment of the quantities required each month and the adaptation of the system to variations in demand. The tool was developed in Excel. On the first sheet, instructions were left, along with possible errors that may occur (Figure 7) to assist the users, who are the supervisors of each manufacturing sector. Instructions and possible errors that may arise when updating the file to add an SKU to a manufacturing sector or add a new manufacturing sector have also been included, considering the future implementation of the Kanban system in other manufacturing sectors.
The first step in using the Excel tool is to gather all accepted orders for the month and conduct an availability analysis to translate these orders into a list of all required SFPs and their quantities. This step does not represent extra work for the supervisor, as there is already a file shared monthly with this information. Next, the Excel tool has an 'ABC Analysis' button that performs ABC analyses of the various manufacturing sectors in the file, thus suggesting the SKUs that should be included in the Kanban system. Since the number of teams and shifts working on each product family on the assembly lines may vary each month according to orders and scheduled deliveries of purchased components, the user must update the file with this information. Following these instructions, the output is a list of SKUs by manufacturing sector, ordered by their ABC classification. The Excel tool indicates how many Kanban cards are needed to satisfy the demand on the assembly lines, considering the maximum production rate. The SKUs already included in the Kanban system are identified in blue (Figure 8).
Much of the data required to prepare this file has already been collected. However, the columns of 'next code' and its respective 'sector' were added to assist in the evaluation of each SKU, considering, for example, the production conditions of the sector, such as the existence of long setup times that require larger batches, and whether the team has already received training to work with the Kanban system. In addition to the sectors where the Kanban system is already implemented (i.e., LPB sector), additional information was collected to create spreadsheets for other manufacturing sectors, leaving the information processed to allow the expansion of the Kanban system.
4.3 Results
The main objective of the work was to reduce supply failures on the assembly lines and the downtime accumulated due to these failures. BPMN was used to improve the planning process, and Kanban was used at the production control level to achieve this goal. The joint application of these tools allowed the reduction of accumulated downtime on assembly lines, thus supporting the effectiveness of their application. Comparing weekly SFPs supply failures from October 2020 to April 2021 shows that the maximum downtime value decreased from about 3,500 minutes (58 hours) before the improvements were implemented to about 2,000 minutes (33 hours) after the improvements were implemented (Figure 9).
On the one hand, modeling the sequencer creation process delivered to the LPB sector with BPMN allowed for the identification of opportunities to improve communication between the Planning and Production departments, eliminating the existence of several sources of information for the same sector, which were frequently contradictory. In this way, a clear information flow was created and adapted to the sector's needs. The sequencer is now delivered quicker, providing greater flexibility for plan changes. Furthermore, defining a minimum order quantity improved coordination between the Production and Procurement departments. On the other hand, implementing the Kanban system brought advantages beyond reducing supply failures on assembly lines, such as creating fixed locations for placing stock next to the manufacturing sectors. In this way, the time and distance travelled in search of SFPs has been reduced, and the creation of fixed supply routes in the future is now possible. In addition to changes in location, the Kanban system also allowed for a reduction in quantities and respective capital invested in intermediate stock. By the end of 2020, inventory levels were rising at the factory. However, with the implementation of the Kanban system, inventory levels began to decrease in all manufacturing sectors.
In the first month after implementing the Kanban system, the inventory levels decreased from 15,584 to 10,133 units in the laser machine sector, representing a decrease of approximately 35%. In panel bender machines, inventory levels decreased from 3,241 to 1,927 units, representing a decrease of approximately 40%. For bending machines, values decreased by 20% from 6,089 to 4,896 units.
Financially, the intermediate stock represented an investment of around €36,650, but after implementing the Kanban system, the amount decreased by around 29% to €26,050. Considering storage costs, these savings represent an effective gain for the company.
5 Discussion and conclusion
This article presents the case of a manufacturing company that aimed to improve the production planning and control of SFPs. The study enriches the literature on LM, presenting a case study where the Lean approach is complemented with BPMN. As far as is known, this is the first case in which the application of BPMN is combined with a Kanban system, exploring the potential of complementing LM with other philosophies that also aim at process improvement. BPMN is an alternative to VSM (widely used in LM implementation), as it is better suited to analyzing information flows and tasks affecting the various stakeholders in the process. Additionally, LM literature has mainly focused on presenting case studies on the application of less advanced tools and lacks examples of more advanced tools, such as the Kanban system. According to Badhotiya et al. (2024), only ten case studies have been published in the last decade addressing the creation of pull systems. This study offers a detailed description of the implementation of a Kanban system, presenting effective results of its implementation on the shop floor, unlike other studies that present results using simulation and the design of the future VSM (e.g., Sharma et al., 2022; Wang et al., 2024). The presented case study also considers the dynamic nature of demand, which is typically not considered in more traditional Kanban systems.
The description of this case study contributes to the creation of critical mass for the definition of a roadmap for the implementation of LM, considering the integration with other philosophies, which can be transversal to many organizations, making it less dependent on the context. Companies with similar characteristics or problems can learn several lessons from this example. The case highlights the role that BPMN process modeling can play in identifying improvements in information flows related to the intervened processes, something scarcely explored in the context of manufacturing companies (Calçado et al., 2024). In this case, improving the planning process was essential because the output of the planning process is used as an input to the Kanban system. The study highlights the importance of involving employees as much as possible in the process and adapting the system to existing procedures without ignoring the need to eliminate waste, contributing to creating a culture of participation where employees themselves identify problems and propose improvements based on their experience. Despite the growing awareness about the importance of leadership involvement in LM implementation, adopting a bottom-up approach continues to be the basis of the improvement process (Maldonado et al., 2020). Thus, this case study demonstrates that it is necessary to train not only the operators who will be the primary users of the Kanban system, but also all participants in the upstream and downstream sectors who interact with the sector where the intervention occurred. Following a standardized approach helps guide the implementation process; however, adaptations may be necessary considering the application particularities of the context (e.g., available space) or the characteristics of the process. Finally, this case study highlights the importance of defining clear rules and audit procedures to ensure system maintainability.
Because it was conducted in a specific context, the results and conclusions of this work may not be generalizable. Still, they demonstrate the applicability of BPMN and Kanban to improve the production planning and control process, with effective gains for the company studied. It will be important to apply the same type of solutions to other cases and contexts to reinforce the impact of this type of implementation.
Statement on Data Availability
All relevant data are included in the article. No additional data will be provided due to confidentiality concerns.
Funding
How to cite:
References
-
Badhotiya, G. K., Gurumurthy, A., Marawar, Y., & Soni, G. (2024). Lean manufacturing in the last decade: insights from published case studies. Journal of Manufacturing Technology Management, 35(4), 766-798. http://doi.org/10.1108/JMTM-11-2021-0467
» http://doi.org/10.1108/JMTM-11-2021-0467 -
Calçado, R., Ávila, L., & Rosa, M. J. (2024). Combining business process management and lean manufacturing to improve information and documentation flows: a case study. Business Process Management Journal, 30(7), 2564-2585. http://doi.org/10.1108/BPMJ-10-2023-0773
» http://doi.org/10.1108/BPMJ-10-2023-0773 -
Chan, F. T. S. (2001). Effect of kanban size on just-in-time manufacturing systems. Journal of Materials Processing Technology, 116(2–3), 146-160. http://doi.org/10.1016/S0924-0136(01)01022-6
» http://doi.org/10.1016/S0924-0136(01)01022-6 -
Chinosi, M., & Trombetta, A. (2012). BPMN: an introduction to the standard. Computer Standards & Interfaces, 34(1), 124-134. http://doi.org/10.1016/j.csi.2011.06.002
» http://doi.org/10.1016/j.csi.2011.06.002 -
Dani, V. S., Dal Sasso Freitas, C. M., & Thom, L. H. (2019). Ten years of visualization of business process models: a systematic literature review. Computer Standards & Interfaces, 66, 103347. http://doi.org/10.1016/j.csi.2019.04.006
» http://doi.org/10.1016/j.csi.2019.04.006 -
Dumas, M., La Rosa, M., Mendling, J., & Reijers, H. A. (2018). Fundamentals of business process management. Berlin: Springer. http://doi.org/10.1007/978-3-662-56509-4
» http://doi.org/10.1007/978-3-662-56509-4 -
Ferreira, G. S. A., Silva, U. R., Costa, A. L., & Pádua, S. I. D. D. (2018). The promotion of BPM and lean in the health sector: main results. Business Process Management Journal, 24(2), 400-424. http://doi.org/10.1108/BPMJ-06-2016-0115
» http://doi.org/10.1108/BPMJ-06-2016-0115 - Gross, J., & McInnis, K. (2003). Kanban made simple: demystifying and applying Toyota’s legendary manufacturing process New York: Amacom.
-
Guneri, A. F., Kuzu, A., & Taskin Gumus, A. (2009). Flexible kanbans to enhance volume flexibility in a JIT environment: a simulation based comparison via ANNs. International Journal of Production Research, 47(24), 6807-6819. http://doi.org/10.1080/00207540802425351
» http://doi.org/10.1080/00207540802425351 - Kitchenham, B., Pickard, L., & Pfleeger, S. L. (1995). Case studies for method and tool evaluation. IEEE Software, 12(4), 52-62.
-
Maldonado, M. U., Leusin, M. E., Bernardes, T. C. A., & Vaz, C. R. (2020). Similarities and differences between business process management and lean management. Business Process Management Journal, 26(7), 1807-1831. http://doi.org/10.1108/BPMJ-09-2019-0368
» http://doi.org/10.1108/BPMJ-09-2019-0368 -
Mamoojee-Khatib, H., Antony, J., Teeroovengadum, V., Garza-Reyes, J. A., Tortorella, G. L., Foster, M., & Cudney, E. A. (2023). A systematic review of lean implementation frameworks and roadmaps: lessons learned and the way forward. The TQM Journal http://doi.org/10.1108/TQM-09-2023-0280
» http://doi.org/10.1108/TQM-09-2023-0280 -
Maware, C., Okwu, M. O., & Adetunji, O. (2022). A systematic literature review of lean manufacturing implementation in manufacturing-based sectors of the developing and developed countries. International Journal of Lean Six Sigma, 13(3), 521-556. http://doi.org/10.1108/IJLSS-12-2020-0223
» http://doi.org/10.1108/IJLSS-12-2020-0223 -
Naeemah, A. J., & Wong, K. Y. (2023). Selection methods of lean management tools: a review. International Journal of Productivity and Performance Management, 72(4), 1077-1110. http://doi.org/10.1108/IJPPM-04-2021-0198
» http://doi.org/10.1108/IJPPM-04-2021-0198 -
Negrão, L. L. L., Godinho, M., Fo., & Marodin, G. (2017). Lean practices and their effect on performance: a literature review. Production Planning and Control, 28(1), 33-56. http://doi.org/10.1080/09537287.2016.1231853
» http://doi.org/10.1080/09537287.2016.1231853 - Ohno, T. (1988). Toyota production system: beyond large-scale production New York: Productivity Press.
-
Rahman, N. A. A., Sharif, S. M., & Esa, M. M. (2013). Lean manufacturing case study with kanban system implementation. Procedia Economics and Finance, 7, 174-180. http://doi.org/10.1016/S2212-5671(13)00232-3
» http://doi.org/10.1016/S2212-5671(13)00232-3 -
Recker, J. (2010). Opportunities and constraints: the current struggle with BPMN. Business Process Management Journal, 16(1), 181-201. http://doi.org/10.1108/14637151011018001
» http://doi.org/10.1108/14637151011018001 -
Reddy, K. J. M., Rao, A. N., Lanka, K., & Gopal, P. (2023). System dynamics modelling of fixed and dynamic Kanban controlled production systems: a supply chain perspective. Journal of Modelling in Management, 18(1), 17-35. http://doi.org/10.1108/JM2-06-2020-0168
» http://doi.org/10.1108/JM2-06-2020-0168 -
Santhiapillai, F. P., & Ratnayake, R. M. C. (2024). BPMN-VSM-based knowledge-work process analysis and improvement in police services: a case study. International Journal of Productivity and Performance Management, 73(7), 2116-2135. http://doi.org/10.1108/IJPPM-12-2022-0604
» http://doi.org/10.1108/IJPPM-12-2022-0604 -
Sharma, A., Bhanot, N., Gupta, A., & Trehan, R. (2022). Application of Lean Six Sigma framework for improving manufacturing efficiency: a case study in Indian context. International Journal of Productivity and Performance Management, 71(5), 1561-1589. http://doi.org/10.1108/IJPPM-05-2020-0223
» http://doi.org/10.1108/IJPPM-05-2020-0223 -
Soliman, M., Oliveira Esteves, O., Trevisan, M., & Fogliarini Segatto, G. (2022). A tentative integration of value stream mapping (VSM) and BPMN for improved process mapping. Knowledge and Process Management, 29(4), 371-382. http://doi.org/10.1002/kpm.1729
» http://doi.org/10.1002/kpm.1729 -
Thürer, M., Fernandes, N. O., & Stevenson, M. (2022). Production planning and control in multi-stage assembly systems: an assessment of Kanban, MRP, OPT (DBR) and DDMRP by simulation. International Journal of Production Research, 60(3), 1036-1050. http://doi.org/10.1080/00207543.2020.1849847
» http://doi.org/10.1080/00207543.2020.1849847 -
Tortorella, G. L., Marodin, G. A., Miorando, R., & Seidel, A. (2015). The impact of contextual variables on learning organization in firms that are implementing lean: a study in Southern Brazil. International Journal of Advanced Manufacturing Technology, 78(9-12), 1879-1892. http://doi.org/10.1007/s00170-015-6791-1
» http://doi.org/10.1007/s00170-015-6791-1 -
Voss, C., Tsikriktsis, N., & Frohlich, M. (2002). Case research in operations management. International Journal of Operations & Production Management, 22(2), 195-219. http://doi.org/10.1108/01443570210414329
» http://doi.org/10.1108/01443570210414329 -
Wang, C. N., Vo, T. T. B. C., Chung, Y. C., Amer, Y., & Truc Doan, L. T. (2024). Improvement of manufacturing process based on value stream mapping: a case study. EMJ - Engineering Management Journal, 36(3), 300-318. http://doi.org/10.1080/10429247.2023.2265793
» http://doi.org/10.1080/10429247.2023.2265793 - Womack, J., & Jones, D. T. (2003). Lean thinking: banish waste and create wealth in your corporation (2nd ed.). Washington: Free Press.
-
Yazdani, A. A., & Tavakkoli-Moghaddam, R. (2012). Integration of the fish bone diagram, brainstorming, and AHP method for problem solving and decision making-a case study. International Journal of Advanced Manufacturing Technology, 63(5–8), 651-657. http://doi.org/10.1007/s00170-012-3916-7
» http://doi.org/10.1007/s00170-012-3916-7 - Yin, R. (1994). Case study research : design and methods (2nd ed). Thousand Oaks: Sage Publications.


















