Abstract
Purpose
Pursuing digital transformation is a valuable strategy to attain companies’ operational excellence and sustainable development. However, installing digital technologies and software is insufficient for achieving a successful digital transformation. Equally vital is providing digital solutions with reliable input data, which is a hard task in companies where information is gathered through manual or non-standardized processes. The lack of reliable data prevents technologies and software from operating at their best, hindering their ability to process information and derive correct insights for improvement. To avoid this, companies should embrace structured problem-solving approaches to evaluate current data retrieval processes, identify error sources and formulate countermeasures. This paper aims to provide an empirical study to substantiate A3 as a winning approach for advancing input data acquisition in companies.
Design/methodology/approach
A case study research is proposed, investigating the application of A3 in an ink manufacturing company, and checking how A3 improves data collection and company performance.
Findings
The case study corroborates A3 as an effective approach, allowing the removal of inefficiencies that previously went unnoticed and reaching a 10% productivity improvement.
Practical implications
Pursuing digital transformation is a valuable strategy to attain companies’ operational excellence and sustainable development. However, installing digital technologies and software is insufficient for achieving a successful digital transformation. Equally vital is providing digital solutions with reliable input data, which is a hard task in companies where information is gathered through manual or non-standardized processes.
Originality/value
The originality of this paper lies in the application scope and aim of the A3 approach. A3 is a mature paradigm but often coined for production and still endowed with unexplored potential. This paper proposes the application of A3 for improving companies’ data retrieval processes, focusing for the first time on information reliability and its importance in ensuring the functioning of digital technologies and software.
Keywords
Citation
Cantini, A., Costa, F. and Portioli-Staudacher, A. (2024), "Driving corporate digitization with reliable data through the A3 approach: an Italian case study", International Journal of Lean Six Sigma, Vol. 15 No. 8, pp. 143-170. https://doi.org/10.1108/IJLSS-03-2024-0055
Publisher
:Emerald Publishing Limited
Copyright © 2024, Alessandra Cantini, Federica Costa and Alberto Portioli-Staudacher.
License
Published by Emerald Publishing Limited. This article is published under the Creative Commons Attribution (CC BY 4.0) licence. Anyone may reproduce, distribute, translate and create derivative works of this article (for both commercial and non-commercial purposes), subject to full attribution to the original publication and authors. The full terms of this licence may be seen at http://creativecommons.org/licences/by/4.0/legalcode
1. Introduction
The era of globalized manufacturing confronts many challenges, encompassing rapidly changing customer demands, fierce market competition, the imperative for short delivery times and heightened requirements for innovative products at reduced costs (Yin et al., 2018). In this context, the pursuit of operational excellence and sustainable development becomes crucial for companies’ survival (Roche and Baumgartner, 2024). It allows for providing quality products and services at affordable costs, while ensuring customer satisfaction and reducing the environmental and social impacts of manufacturing processes (Henriquez et al., 2023).
A valuable strategy for achieving operational excellence and sustainable development in industries involves adopting Industry 4.0 (I4.0) technologies and software in manufacturing systems (Ferraro et al., 2023; Virmani and Ravindra Salve, 2023). Notably, augmented reality, artificial intelligence, robots and other digital solutions enable the automated and efficient application of continuous improvement methodologies (Muhuri et al., 2019). Indeed, these technologies can uncover hidden patterns and meanings in input data, allowing for data-driven decision-making to improve manufacturing processes under operational and sustainability aspects (AlKhader et al., 2023; Fiorello et al., 2023).
In this scenario, many companies are incorporating digital solutions into their operations, witnessing a sharp growth in the installation of I4.0 technologies and software (Li et al., 2019). For instance, many small and medium enterprises have recognized digital transformation as a primary strategy for remaining competitive in the market (Mishrif and Khan, 2023). However, solely installing new technologies and software is insufficient to attain operational excellence and sustainable development (Yusof et al., 2012). Equally vital is providing digital solutions with reliable input data. This task poses challenges in companies where the acquisition of data is manual or follows non-standardized processes (Torri et al., 2021). The absence of the right input data in the right place and timing deprives technologies and software from operating at their best, hindering their ability to process information and derive insights for improvement (Koreček et al., 2020). Consequently, companies experience two side effects. First, technologies and software may conduct erroneous analyses, proposing inappropriate improvement interventions and potentially causing declines in performance (rather than enhancements). Second, unsuccessful enhancements in operational and sustainability performance impede remunerating the economic investments required for installing technologies and software, posing issues when financial resources are constrained (AlKoliby et al., 2023).
To avoid these side effects, companies should embrace structured problem-solving approaches to evaluate current data retrieval processes, identify sources of error and formulate countermeasures (Battistella et al., 2023; Rossini et al., 2021). A powerful problem-solving approach for this purpose is the “A3” approach offered by the Lean philosophy (Shook, 2008). A3 aims to reinforce the application of the Plan-Do-Check-Act (PDCA) cycle through an intuitive and visual structure called the “A3 report.” A3 emerges as an ally to pursue operational excellence and sustainability in companies, owing to its simplicity and cost-effectiveness (Santhiapillai and Ratnayake, 2023). These characteristics, as noted by Nguyen (2015), facilitate the initiation of continuous improvement (Kaizen) processes, even in companies with financial constraints and no Lean experts. Embracing A3 will empower companies with a structured problem-solving approach to enhance the reliability of input data acquisition, fully harnessing the capabilities of digital technologies and software to attain operational excellence and sustainable development (Santos Filho and Simão, 2022).
While A3 has proven to be effective in enhancing operational and sustainability performance across various contexts (cf. Section 2), the literature reveals few A3 applications targeting the improvement of data management processes within companies (Kassem et al., 2023; Torri et al., 2021). Specifically, to the best of the authors’ knowledge, no study has harnessed A3 to ensure the provision of reliable input data for feeding digital technologies and software, thereby facilitating successful transitions to smart industries. This gap deprives companies of awareness of a structured (yet simple and cost-effective) approach to enhance their innovation capability, leverage the potential of I4.0 solutions and strive for operational excellence and sustainable development (Bevilacqua et al., 2017). To fill this gap, this paper aims to address the following research question:
How can A3 be used to improve data input and retrieval in digitalized companies?
Accordingly, the purpose of this paper is to present an empirical study to substantiate A3 as a winning approach for advancing data acquisition in companies. Particularly, this paper proposes the application of A3 in the case study of an ink manufacturing company. The remainder of this paper is as follows. Section 2 presents the background literature on A3 applications in information management contexts. Section 3 outlines the methodology followed in this paper, deepening the A3 approach. Section 4 introduces the case study. Section 5 presents the case study results. Section 6, discusses the achieved results. Finally, Section 7 offers concluding remarks.
2. Theoretical background
A3 is a structured problem-solving approach to carry out Kaizen projects (Ferrazzi et al., 2023). Developed by Sobek II and Jimmerson (2004), A3 is rooted in the principles of the Toyota Production System, providing a systematic approach for contemplating new ways to improve. A3 reinforces the application of the PDCA cycle, involving understanding and step-by-step addressing a problem by documenting it on an A3-sized paper: the A3 report (Santhiapillai and Ratnayake, 2023). While details on the A3 report’s content will be provided in Section 3, it is worth mentioning that the strength of the A3 approach lies in its role as a guide for project teams (Low et al., 2022). A3 aids in identifying inefficiencies, making collective decisions and arriving at solutions through consensus among team members (Naik et al., 2012). By writing the A3 report, an objective communication tool will be shared and reviewed by cross-departmental stakeholders, fostering collaborative problem-solving while tracking the progression of the resolution process.
According to Sobek II and Jimmerson (2004), the first case study applications of A3 have been made in Toyota Motor Corporation, to reduce production waste while increasing quality. Besides the automotive sector (Kanyinda et al., 2020), A3 has also been applied in other industrial sectors, including the aeronautical (Gnanaguru et al., 2011), metal (Apaza-Casabona et al., 2022), energy (Al Busaidi and Al-Busaidi, 2021), food (Marques et al., 2022) and so on. A3 has not only been proposed to improve plants’ performance but also humans (Zuñiga et al., 2023), sustainability (Fuenzalida et al., 2016) and supply chain performance (Rojas et al., 2023). Moving from manufacturing to service organizations, A3 has been applied to empower pedagogical capabilities in education (Wang et al., 2019), enhance police districts’ performance (Santhiapillai and Ratnayake, 2021) and optimize healthcare systems. The healthcare system is the most investigated service sector (Sobek and Jimmerson, 2006), showing several A3 applications, such as in Naik et al. (2012) and Akbulut et al. (2023).
Beyond the aforementioned applications, recent studies have highlighted A3 as a facilitating tool for integrating I4.0 technologies and software within companies. For example, Frecassetti et al. (2023) leveraged A3 to introduce automated guided vehicles in warehouses. However, as already discussed, merely installing digital solutions in companies is not sufficient to ensure their functionality; they also need to be fueled with reliable data. Although the use of A3 is quite mature, according to Kassem et al. (2023) and Torri et al. (2021), its literature applications are often coined for production, leaving unexplored potentials in the field of Information Technology (IT) and data management. The studies in this domain include Singh and Muller (2013), Krogstie et al. (2014), Khodambashi (2015) and Carpentari et al. (2020). These works showcase how A3 can enhance documentation, communication and knowledge-sharing processes among departments and operators. However, the extant literature primarily focuses on streamlining company information flows, aiming to minimize exchanges of documents, with the assumption that shared data is reliable. Yet, to the best of the authors’ knowledge, none of these studies adopting A3 has delved into the impact of unreliable information on the successful implementation of digital technologies and software.
Concerning other studies on Lean philosophy, recent literature has shown an emerging research stream investigating how to conduct Kaizen processes in companies embarking on paths of digitalization. Examples of studies belonging to this stream include Powell (2024), Terelak-Tymczyna and Niesterowicz (2024) and Tsukada et al. (2024), which explore the introduction of artificial intelligence, blockchain, augmented reality and other I4.0 technologies into different manufacturing processes. Chivukula and Pattanaik (2023) have also prioritized various groups of Lean tools to support digital transitions by evaluating their compatibility with I4.0 technologies. In addition, Frecassetti et al. (2024) have discussed how to leverage Lean practices to overcome barriers in companies implementing digitalization. However, these studies, while exploring several Lean tools (e.g. value stream mapping, Kanban, Heijunka, etc.) to tackle the digital transformation challenge, have not yet proposed the use of A3 for improving data input and retrieval in digitalized companies. This paper addresses this gap by demonstrating that A3 can serve as a valuable problem-solving approach to ensure the reliability of input data for technologies and software, thereby facilitating transitions to I4.0 and contributing to higher levels of operational excellence and sustainability.
3. Methodology
The research methodology used in this paper is a single instrumental case study (Yin, 2018). Following Karlsson et al. (2023), a case study is analyzed to test and confirm an a priori theory-based hypothesis that drives the research, rather than to formulate a new theory. Specifically, the case study herein presented aims to confirm the theoretical effectiveness of A3 as a structured problem-solving approach to evaluate data retrieval processes in companies with digital technologies and software, identify sources of error and develop countermeasures. The choice to use a case study research methodology is grounded in several considerations:
A case study enriches the knowledge of researchers and practitioners by corroborating theoretical concepts through practical examples (Eisenhardt and Graebner, 2007). Witnessing the successful application of a theory in real-world scenarios encourages researchers and practitioners to use the same theory to address similar issues within their companies.
A case study allows for investigating a phenomenon within its natural setting, which is essential in information-rich contexts influenced by numerous variables (Gijo et al., 2018). Indeed, a case study allows for working in the field, making direct observations and engaging with the company employees. This direct involvement is crucial for collecting data in natural settings and uncovering the authentic challenges faced by companies (Cantini et al., 2020; Voss, 2008).
A case study provides versatility in design and application, as it allows both qualitative and quantitative analyses (Sunder M et al., 2020).
Finally, the case study methodology has proven particularly valid in analyzing the effectiveness of Lean tools such as the A3 report (Rojas et al., 2023).
Even if the findings of a single case study can be generalized to a limited extent, Sunder M et al. (2020) claim that each case study adds to the body of knowledge by offering a deeper comprehension of the investigated phenomenon.
Regarding the approach adopted to carry out the case study, aligning with the aim of this paper we adopted the A3 approach. As said, the A3 approach entails applying the PDCA cycle supported by an intuitive and visual structure: the A3 report. Figure 1 outlines the A3 report, which is composed of eight sections each reflecting a specific step of the PDCA cycle.
The names of the sections in Figure 1 may change, but the basic storyline behind the A3 report remains the same. The left-hand sections in Figure 1 show the current company situation, investigating causes of inefficiency and possible countermeasures. Whereas the right-hand sections show the improved process (after implementing countermeasures). The sections must be travelled from left to right, following the arrows. When using the A3 approach, each section of the A3 report has to be completed with the information described by Sobek II and Jimmerson (2004).
First, the addressed problem (slated for improvement) must be elucidated. This involves stating the problem and describing any background information essential for understanding its importance. Consequently, in Section 1 of the A3 report (i.e. clarify the problem), the information to be incorporated encompasses what is the addressed issue, how it was discovered and why it is relevant to the company’s goals. As recommended by Kanyinda et al. (2020), to fill in Section 1 of the A3 report, it is advisable to consult the company board and operators, interviewing them about the problem affecting their company. Indeed, the A3 approach always begins by defining the main issue through the eyes of the customer.
Second, the problem symptoms are delineated, including figures, tables or sentences within Section 2 of the A3 report (i.e. breakdown the problem) to prove how the current business performance results in the identified problem. Accordingly, to fill in this section of the A3 report, any company database should be consulted, data-driven analyses conducted, and insights from the field extracted to better understand the current situation and provide clear evidence of the problem defined in the previous section.
Third, agreeing with the company board, the expected targets to be reached with the A3 implementation are established, articulating them within Section 3 of the A3 report (i.e. set the target). Therefore, filling in this section involves listing the final goals to be achieved, eventually categorizing them into must-have (essential) and nice-to-have (desirable) targets.
Fourth, a root cause analysis is conducted to discern the roots of the identified problem. By using tools like the Ishikawa diagram or the 5-Whys, the identified causes are reported within Section 4 of the A3 report (i.e. analyze the root causes). Accordingly, filling in Section 4 entails outlining the original causes of the problem, thus laying the foundation for then removing the roots of the issue rather than simply treating its superficial symptoms.
Fifth, building upon the identified root causes, potential countermeasures are proposed to solve the problem and achieve the predefined target. These countermeasures are summarized in Section 5 of the A3 report (i.e. develop countermeasures). They can be conceived by organizing brainstorming sessions focused on identifying respective solutions for each root cause.
Sixth, meetings are organized with the company board to deliberate on the proposed countermeasures and formulate implementation plans for all of them or the most promising ones. Hence, Section 6 of the A3 report (i.e. implement countermeasures) encompasses details on the necessary activities for countermeasure implementation, the timeline and modalities of implementation, and the designated individuals responsible for the execution. Using a Gantt chart can be beneficial for summarizing this information.
Seventh, the advancement of activities required for implementing countermeasures is monitored. This ensures compliance with the countermeasures’ implementation plan and assesses whether the activities are producing the expected impacts (for reaching the target). Therefore, in the seventh section of the A3 report (i.e. monitor results), information is provided on how and when monitoring activities will occur, progressively reporting the follow-ups. When presenting the follow-ups, the initial company situation (as outlined in Section 2 of the A3 report) can be compared with the situation after implementing the countermeasures. This comparison enables the verification of results and the confirmation of the positive effects produced by the A3 approach.
Finally, to perpetuate the application of the PDCA cycle, the lessons learned are documented and shared, establishing the groundwork for subsequent improvement initiatives. Accordingly, in the eighth section of the A3 report (i.e. standardize and share success), information is provided regarding the procedures to guarantee and sustain the success achieved through the current Kaizen process. In addition, information is provided on how the lessons learned will be shared with other stakeholders.
Overall, the A3 approach (supported by the creation of the A3 report) empowers companies with a structured problem-solving approach to engage in Kaizen endeavors.
4. Case study description
Company A is an ink manufacturer for multiple applications: coating, printing, etc. It has 200 years of expertise and takes pride in producing tailored inks to meet individual customer needs. The ad hoc products vary in terms of delivered quantity, hue and chemical composition (e.g. organic, metallic, etc.). However, inks are all composed of three raw materials: pigment (that ensures the color and represents 25% of the product), resin (that ensures the adhesion and represents 25% of the product) and solvent (that allows inks to dry and represent 50% of the product).
In the considered Italian plant of Company A, the inks are produced on several production lines, each dedicated to a specific color (e.g. yellow, red, etc.). Each line produces inks by applying the five sequential phases (Figure 2).
In the loading phase, an operator retrieves the ink recipe from the IT system, loading the pre-dispersion machine with the required solvent, resin and pigment. These latter undergo the pre-dispersion phase, being stirred for more than 1 h to ensure product homogeneity. The pre-dispersed product is transferred to Tank 1, while the pre-dispersion machine is cleaned to allow the next batch production. Cleaning involves automatically adding a small quantity of solvent within the pre-dispersion machine, recirculating it for some minutes and transferring it to Tank 1. After cleaning, milling 1 is performed to strengthen the ink color. To this end, the pre-dispersed product is transferred from Tank 1 into a vertical bead mill, where beads located between a stator and a rotor crush the pigment particles. After milling 1, a semi-finished product called “base” is obtained and transferred to Tank 2 (waiting for milling 2). As in the pre-dispersion phase, the milling 1 machine and Tank 1 are cleaned with solvent. Next, milling 2 starts, which is similar to milling 1 but adopts a different rotor speed and beads with a lower size to further refine the pigment particles. Now, the base is transferred to Tank 3, while the milling 2 machine and Tank 2 are cleaned. Manual quality control is performed on Tank 3, adjusting the base recipe by adding solvent or resin if needed. Finally, the potting phase is performed, which involves filling different-size containers according to customer requests. During potting, the base is handled by an operator who is responsible for weight inspections and labeling. The five production phases in Figure 2 follow a make-to-stock policy. The resulting bases are stored in a warehouse of semi-finished products. The final ink is produced according to a make-to-order (MTO) policy by mixing two or more bases in varying proportions. MTO enables product customization while minimizing inventory levels, allowing for more than 2,000 colors to be produced with 100 bases. However, the MTO production occurs in a separate plant (not the one considered in this paper). Therefore, it is excluded from Figure 2, limiting this analysis to the production of bases.
Company A strives for operational excellence and sustainable development to uphold customer satisfaction and maintain market leadership. Recognizing the pivotal role of technologies and software in conveying operational excellence and sustainability, Company A started a massive digitalization program. Accordingly, the Italian production plant was equipped with cutting-edge solutions to track production activities and record data emanating from machines. Moreover, Company A invested in software to monitor the Overall Equipment Effectiveness (OEE) of production lines, aiming to automate the derivation of data-driven insights to minimize waste and pinpoint opportunities for improvement. However, Company A is encountering challenges in integrating the OEE software into its daily use due to issues with the input data retrieval.
5. Case study results
To delve into Company A’s problem and find solutions, the A3 approach was applied. The results are detailed below, providing a paragraph for each section of the A3 report (Figure 1). Appendix 1, then, provides the overall A3 report.
5.1 Clarify the problem
Implementing the A3 started by conducting meetings and interviews for exploratory purposes. Two meetings were held with the company board to grasp the company’s culture and background information. It emerged that Company A’s culture is oriented toward meeting customer needs while ensuring reliability, on-time delivery, product quality and customized products. In addition, face-to-face interviews were carried out with project managers, IT operators and production operators to clarify the problem encountered with the OEE software. It emerged that the OEE software represents an ally to pursue the company’s culture but the problem experienced can be stated as follows. The OEE software is unreliable and remains unused due to issues in defining reference parameters and retrieving input data for the OEE calculation. The lack of the right input data in the right place and time hampers Company A to detect production inefficiencies and implement data-driven Kaizen improvements.
5.2 Breakdown the problem
To confirm this problem and elucidate Company A’s current situation, Gemba walks (direct observations of processes) were conducted, accompanied by additional interviews with company employees. As a result, the current data retrieval and exchange process was outlined, resulting in the flow in Figure 3.
In Figure 3, an operator consults the enterprise resource planning (ERP). Based on the workload schedule, the ERP informs the operator about the designated batch (order number) to process. Accordingly, the operator extracts from the ERP the recipe of the base to be produced. Then, the operator enters into the production software the ERP information, namely the produced base, the batch size, and the starting production time. Consistently, the operator loads raw materials in the line (see Figure 2), after which the production automatically starts dictated by the production software. The line performance is real-time recorded by sensors, storing in a database variables such as temperature, energy, power and flows. Before installing the OEE software, the data recorded by sensors was solely used for monitoring the process (ensuring that it was under control). To leverage the recorded data for aggregated analyses, Company A has recently invested in the OEE software. Upon receiving appropriate input data, the OEE software enables Company A to measure the current production efficiency compared to its full potential. Therefore, inefficiencies and opportunities for improvement in each production line can be identified as follows. The OEE is the product between three factors [equation (1)]: availability (A), capacity (C) and quality (Q):
Availability [A, equation (2)] measures the amount of time the line is operational. It is the ratio between uptime (i.e. effective working time) and the net available time (NAT, namely the theoretical time available for production). Uptime differs from NAT due to standstill time (ST), which is the sum of downtimes including stops, setups, etc. Capacity [C, equation (3)] gauges the production speed relative to its designed speed. It is the ratio between the actual output produced (expressed in terms of production time) and the uptime. The output produced, in turn, is the product of the ideal cycle time for producing one item (ICT) and the total number of items produced (TI). Quality [Q, equation (4)] measures the percentage of produced items that meet the quality standards. It is the ratio between compliant items (CI) and the total number of items produced. In equation (1), the ideal OEE value is 1. Any deviation from unity highlights losses related to downtimes, speed reductions, or non-compliance. By continuously updating the OEE calculation and generating performance reports, the OEE software provides Company A with insights into inefficiencies and opportunities for improvements.
However, to work, the OEE software must be fed with input data on equipment working times, the number of items produced, their compliance, etc. Currently, feeding the OEE software with input data is neither an automated nor a standardized procedure (see Figure 3). The IT department of Company A manually extracts data from the production software, loading it as input for the OEE software. In addition, for each production line, a constant reference parameter (i.e. ICT) is needed to calculate the OEE. The ICT value was manually incorporated as input for the OEE software during its initial installation, being established based on employees’ experience. Particularly, Company A assumed an average duration for the production phases in Figure 2 and summed all averages to determine ICT. Upon analyzing historical OEE reports from the first semester of 2022, symptoms emerged regarding the mismatch between the reference parameters configured in the OEE software and the real production process. For instance, referring to a specific base in a particular production phase (details withheld due to non-disclosure agreements), Figure 4 compares the ICT value set by Company A (red curve) and the actual cycle times measured by sensors for different orders (yellow curve). According to Figure 4, the production line appears to outperform expectations, as the yellow curve often remains below the red threshold (meaning that cycle time is lower than its theoretical value). Based on this, the software calculates an OEE exceeding 1, suggesting employees not changing production processes as they are already optimized. This is unrealistic since the considered production line experienced numerous downtimes, slowdowns and instances of non-compliance throughout 2022.
Figure 4 confirmed the problem stated in Section 5.1: the current data retrieval process and the selection of reference parameters are hindering the reliability and the proper functioning of the OEE software. Addressing this problem represents an indispensable condition for maximizing the digital transformation investment made by Company A, allowing the OEE software to realistically represent the performance of production lines and suggest insights for improvement.
5.3 Set the target
After evaluating the problem background, a focus group was organized with the company board to set the target to be reached with the A3 implementation. A roadmap of three targets was developed, dividing them into must-have and nice-to-have ones (Figure 5). The first must-have target involves defining a reliable approach to read the data obtained through sensors and collecting in a database those required to feed the OEE software. This approach must be tested and validated in the field on one production line. As the second must-have target, a reliable approach must be devised to establish appropriate values for the ICT reference parameters needed to calculate the OEE of production lines. In this way, after our presence in Company A, the OEE software will be fed in a reliable way, updating over time the evaluation of lines’ performance and laying the foundation for envisioning improvement directions. Finally, one nice-to-have target was set, namely to use the OEE software’s insights to increase the productivity of one production line by 10% (where productivity is expressed in kilograms of base produced per employee per day, i.e. kg/man*day).
5.4 Analyze the root causes
To discern the problem’s causes and provide ad-hoc countermeasures, a root cause analysis was conducted, whose results are summarized through the Ishikawa diagram in Figure 6.
In Figure 6, the root causes are listed and categorized into five macro-areas, depending on how they are linked with data measurement, man, machine, software or methodology. Figure 6 shows that some causes are duplicated in different macro areas. This is due to their high interdependency. Delving into the data measurement macro-area, Company A has no clear knowledge of machines’ working time since no specific measurements of these times are made over the year. This aspect prevents the OEE calculation and the functioning of the OEE software. In addition, the sensors measuring line performance sometimes have lecture errors, increasing the OEE unreliability. Regarding the man macro-area, line operators believe they have a high knowledge of cycle times associated with bases’ production, despite the high number and variability of producible bases and the continuous changes in production batches. This aspect undermines the reliability of the OEE software because operators suggest wrong ICT values, being convinced that they are correct. Moreover, line operators do not foster the benefits of the OEE software, which seems only useful to visualize real-time the line performance, but not to improve daily production activities. The mistrust in the OEE software makes operators inhibit its use. Regarding the machine macro-area, machines generate fluctuations and variability in the process when they produce a large variety of bases with different quantities and recipes. This hinders OEE calculations and, therefore, the OEE software’s reliability. Regarding the software macro-area, the OEE software calculates unreliable OEE values because it relies on manually transferred input data and reference parameters that are not consistent with real processes (see Figures 3 and 4). Moreover, as per feedback from company employees, the interface of the OEE software is not user-friendly, and the difficulty in navigation makes the deduced performance improvement suggestions unreliable. Regarding the last macro-area, no smart methodology is currently adopted to automatically collect input data from sensors and transfer them to the OEE software. Since the transfer is manual, sometimes input data are lost, making unreliable the OEE calculation. Currently, the real-time data gathered via sensors are stored in compressed files. This necessitates the IT department to individually open these files, locate the information needed as input by the OEE software and extract it appropriately to be fed into the software. Accordingly, manually transferring input data not only makes the OEE software unreliable but also lengthens the time to run the OEE software, hindering its daily use. Overall, Figure 6 shows the interconnected causes that are negatively affecting the OEE software’s reliability. Most root causes are connected with the input data and reference parameters fed into the OEE software. This aspect underscores that implementing advanced technologies and software alone is insufficient for achieving operational excellence and sustainability in companies. Indeed, the effectiveness of digital solutions hinges on the accuracy of their input data.
5.5 Develop countermeasures
To remove the root causes in Figure 6, four potential countermeasures were conceived:
Gemba walks could be performed to improve the knowledge of machines’ working times. By empirically observing the lines, the measured data could be leveraged to estimate reliable OEE reference parameters (i.e. ICT values) associated with different bases and machines. This countermeasure is intuitive and inexpensive to implement. However, given the high variability of bases produced by Company A, its implementation would require many Gemba walks to observe all production variants multiple times (becoming time-consuming). Moreover, relying solely on Gemba walks would not foster the creation of a smart methodology for collecting input data and feeding the OEE software.
Line sensors could be replaced with more accurate ones to avoid errors in data lectures. However, the current sensors are new and their lecture errors are rare, so replacing sensors would be more expensive than effective as a countermeasure.
Instead of merely presenting the OEE value (as a number), an OEE graph could be integrated into the OEE software to enhance the interface’s user-friendliness. The OEE graph could provide Company A with a visual overview of the line performance, illustrating the factor [among A, C and Q in equation (1)] that is least efficient (holding potential for improvement). The OEE graph could increase the perception of Company A operators of the benefits provided by the OEE software. It can offer visual hints for devising Kaizen processes, leading to daily performance enhancements. However, as the input data required to feed the OEE software (and the corresponding OEE graph) is manually provided by the IT department, real-time updating of the OEE graph is unfeasible (huge time and effort would be required to bring its content up to date).
Finally, a smart framework could be formulated to automatically retrieve information from both historical and new databases, and extract input data for the OEE calculation. This countermeasure could enhance a higher knowledge of cycle times, also allowing the automatic determination of reference parameters to run the OEE software. In addition, it could prevent data loss during the manual transfer from the performance database to the OEE software. Although this countermeasure involves navigating through past databases, which could be time-consuming, it could be suitable for not only removing the majority of root causes but also for achieving the must-have targets.
Table 1 summarizes the proposed countermeasures, delineating their contribution toward eliminating root causes. Notably, no countermeasure was devised to reduce the variability in the production process. This decision stems from the recognition that Company A’s competitive advantage lies in its ability to offer a wide portfolio of customized inks (cf. Section 5.1). Therefore, Company A embraces that machines may introduce fluctuations and variability in the process (particularly when producing numerous base formulations with varying quantities and recipes) even if this hinders the OEE software’s reliability.
Due to time and budget constraints, not all countermeasures in Table 1 could be simultaneously implemented in Company A. Following Torri et al. (2021), each countermeasure was classified within the quadrants of an effort-impact matrix, where the effort encompasses the required implementation time and cost while the impact is the benefit expected on targets. Figure 7 shows the resulting matrix.
Following Figure 7, among the proposed countermeasures, we prioritized the implementation of the smart framework. This choice was driven by its perceived suitability for attaining the must-have targets. Moreover, it was deemed to be particularly effective in the long run, considering that Company A might change machines, bases and pigments in the future.
5.6 Implement countermeasures
To apply the selected countermeasures we developed an implementation plan. All activities necessary for achieving the smart framework were detailed. Then, meetings with the company board were organized to deliberate on their sequence, duration and initiation across the timeline. The Gantt chart in Figure 8 outlines the implementation plan, which involves: (i) procuring and cleaning data to feed the smart framework; (ii) developing and validating the smart framework; (iii) leveraging the smart framework to improve the OEE performance of Company A, testing this improvement through a virtual simulation and a design of experiment (DOE); finally, (iv) in case of good results obtained through simulation, test and validate these changes in the field.
5.7 Monitor results
This section shows how each activity of Figure 8 was implemented, reporting its results in a specific sub-section. The advancement of all activities was monitored over time to ensure compliance with the implementation plan. For instance, the column “Progress” in Figure 8 shows the project status detected during a follow-up on April 1, 2022.
5.7.1 Data gathering and cleaning.
As the first implementation activity, we examined the historical company database, which contains data recorded by sensors. Understanding the available information for each machine in the line was crucial for devising a smart data-driven framework for feeding input data into the OEE software. As an example, during this database consultation, we discovered that, for the loading and pre-dispersion machines (see Figure 2), Company A monitors the performance listed in Table 2
weights of the loaded, mixed and unloaded raw materials;
start and end times of stoppages in loading and mixing operations; and
the velocity and pressure of the pump used for mixing raw materials, alongside its energy consumption.
Similar scrutiny was applied to the data collected for other production machines.
Based on the available data, we engaged with Company A employees to understand which variables could significantly influence the duration of each production phase [consequently impacting ICT, ST and OEE in equations (1)–(3)]. Taking again the loading and pre-dispersion phases as an example, operators reported that the primary variables affecting processing time are the weights and pump velocity. Other variables have no discernible impact on production time. For instance, the temperature within Tank 1 does not affect the pre-dispersion duration, as there is little interaction between different raw materials in terms of energy exchange. Similarly, the impactful variables were investigated for other production phases.
After defining the variables needed for calculating the OEE, their data were automatically retrieved from the historical database by developing a VBA script and an Excel spreadsheet. Conversely, other variables measured by sensors – but not affecting the OEE calculation – were excluded from the analysis, thus cleaning the data. Particularly, starting from three different historical spreadsheets provided by Company A (with a total of 800.000 lines and 20 columns referring to one year of production – from April 2021 to April 2022), three respective data frames were achieved:
“Line data frame” – The rows of this data frame provide information on the variables measured by sensors (and impacting the OEE), reporting the time in which the data was recorded. This data frame does not provide information about which order number of the base is produced and at which production phase it is. Data is recorded continuously, providing a row for each minute of every day.
“Activity data frame” – In the rows of this data frame, for each production order, information is provided on the order number, the phase processed (as recorded by the ERP), the start and end time of the phase, and the duration.
“Description data frame” – In the rows of this data frame, for each order number, information is provided on the phase processed (according to the ERP) and its description.
5.7.2 Smart framework development and validation.
To achieve the first must-have target (providing Company A with a smart and reliable framework for supplying input data to the OEE software) the independent data contained in the three aforementioned data frames were linked to each other. This was required to extract the ST, TI and CI values in equations (2) and (3), thereby calculating the OEE.
First, we merged the “activity data frame” and the “description data frame” by matching their respective information based on the common order number. Order number is a unique ID characterizing the rows of the two data frames. Therefore, merging the data frames was possible by implementing the v.look.up function on an Excel spreadsheet. The output (named “dataframe_join_1”) allows Company A to have the following information on each order processed on each line:
order number;
processed base;
start/end time of the production phase and sub-phase (see Table 2);
phase duration; code of the production phase performed (as per the ERP); and
phase description.
However, we were not sure about the reliability of the data coming from the “activity data frame” (particularly the phases’ duration) since such data was retrieved from the historical outcomes of the OEE software. To validate this reliability, we compared the activity duration of the “dataframe_join_1” with the same data in the “line data frame.” Performing this comparison was not immediate since no unique ID (e.g. the order number) was shared between the “line data frame” and the “dataframe_join_1.” Therefore, to merge the respective information in a single file, we applied the v.look.up function based on the start and end time of phases (which were expressed as “yyyy-mm-dd hh:mm:ss”). Merging the data frames allowed having, for each row, information on the order processed, the production phase, consistent timings and durations of phases, and the variables affecting the OEE. The reliability of the obtained information was tested and validated in the field by manually measuring the timings and sequence of activities of one production line and achieving similar results.
By recording the VBA script in a Macro contained in the spreadsheet, Company A was provided with a smart and reliable framework to clean and merge the recorded data frames (stored in the performance database of Figure 3), and automatically feed the OEE software with input data. Accordingly, the first must-have target was achieved. Next, targeting the second must-have target, a rule was introduced within the smart framework to automatically establish the OEE reference parameters of production lines (i.e. ICT values). The rule involves, first, summing, for each order number produced in each line, the duration of production phases (according to the sequence in Figure 2). In this way, the actual cycle time is determined for each order. Second, order numbers are grouped based on the base produced, so that cycle times experienced for different inks are listed in separate spreadsheets. As an example, Table 3 shows the cycle times registered for a specific base manufactured in a production line. Third, outliers (highlighted in italics in Table 3) are removed from the list of each base by filtering out the cycle times outside the range defined by the mean plus or minus three times the standard deviation. Finally, for each base produced in each line, ICT is identified as the minimum (ideal) value among the remaining list, as shown on the bottom part of Table 3.
To validate the reliability of the reference parameters established by the smart framework, as exemplified for a specific base in Figure 9, we compared the ICT value established by the framework (red curve) with the actual cycle times measured by sensors for different orders (yellow curve). Unlike in Figure 4, ICT in Figure 9 is consistent with the line’s performance. In fact, despite the proximity of the yellow and red curves, the yellow curve remains consistently above the red one. Namely, the actual cycle times exceed the ideal one, leading to an OEE of 0.8 (i.e. below 1, unlike in Figure 4), revealing performance losses and consequent chances for improvement.
Similar comparisons to that in Figure 9 were extended to other bases, confirming the reliability of their reference parameters and accomplishing the second must-have target.
5.7.3 Simulation and design of experiment development.
Once reached the must-have targets, the OEE software (fed with the smart framework) was leveraged to meet the nice-to-have ones. Focusing on one production line (i.e. yellow), we aimed to increase its productivity (kg/man*day) by 10%. To do so, the data gathered by the smart framework was exploited as follows. Upon observing an OEE of 0.8, its components [A, C and Q in equation (1)] were examined to identify the one with the lowest value. C emerged as the component with more potential for improvement, showing a value of 0.7. By looking at the duration and sequence of production phases registered for three orders of the same base (1013597705, 1013597706 and 1013597707), C appeared undermined by a process bottleneck: the milling 1 phase (as illustrated in Figure 10). Milling 1 shows a stable duration, which is longer than the predecessor phase. Whereas the upstream pre-dispersion phase, after the first batch (1013597705), must wait for milling 1 to finish, prolonging the cycle times (i.e. yielding to a high ICT).
A simulation and a DOE were conducted to perform a “what-if analysis,” virtually testing different combinations of line input variables and looking for the one that reduces milling 1 duration (enabling a 10% increase in line productivity). Detailing the results of the simulation model and DOE goes beyond the scope of this paper, which is to show how A3 can improve data retrieval processes for feeding digital technologies and software. However, concerning the simulation and DOE, we report the following information. The simulation model was built in consultation with the quality and production managers of Company A, considering that two milling machine input parameters affect cycle times: mill revolutions per minute (RPM) and pump speed. RPM determines how fast the product receives energy from the milling machine. The higher the RPM, the faster the milling phase will be (high product flow rate), with a high energy rate (W*min) and high temperatures exchanged with the product. Whereas the pump speed influences the number of product recirculations in the milling machine. The higher the pump speed, the higher the product flow rate and the energy rate will be (with low temperatures exchanged but high pressure into the system pipelines).
Taking the exchanged energy rate as the simulation output, the DOE was designed to test the combinations of variables listed in Table 4 (performing a full factorial design with two factors – RPM and pump speed – and two replications). For each combination, Table 4 shows the energy rate obtained as the simulation output (along with the values of temperature, flow rate, and pressure for the sake of completeness). The initial line condition was characterized by a milling 1 machine with an average energy rate of 1 kW supplied to the product every 4:15 min (i.e. 235 W/min). Over a year, Company A produces around 890 yellow batches, each requiring an average of 70 kW of power to be produced. Under this condition, the yellow line requires 4,400 h to produce all 890 batches. Therefore, to improve productivity by 10%, the total time should be reduced by 440 h, achievable if the energy rate is set to 1 kW every 3:50 min (i.e. 260 W/min).
Using the “Response optimizer” tool by the Minitab software, setting an energy rate of 260 kW/min as the target, the optimal combination of input variables for achieving the nice-to-have target was identified (Figure 11). Particularly, a pump speed of 53 m/s and RPM of 555 rounds/min were suggested for yielding an expected energy rate of 261 kW/min.
5.7.4 Implementation of improvement actions and validation.
The improvement solution identified in Figure 11 was implemented in Company A, testing it on the field. We run two batches in the yellow line, setting the pump speed at 53 m/s and RPM at 555 rounds/min. Only two batches were tested for time constraints in the project development. However, both batches confirmed that the new setting in the line’s condition improved the energy rate (and therefore the OEE) as per the nice-to-have target.
5.8 Standardize and share success
We showed the results and informed Company A about the achievement of all targets. Moreover, we reported another accomplishment. Manually feeding the OEE software with input data and reference parameters not only produced OEE mistakes and prevented Company A from improving itself, rather it was also time-consuming. A 1-h Gemba walk was required to manually collect the duration of production phases and cycle time of a single order. However, considering, for example, the yellow inks, 18 possible bases are available, and multiple observations (i.e. order numbers) are required for each base. Therefore, estimating the OEE reference parameters through Gemba walks would have been time-consuming. For instance, considering 10 orders observed per base, 180 h (approximately 22 working days) would have been required to estimate the ICT values for all yellow bases. Instead, the proposed smart framework yielded the same results in only 26 h, of which 4 were used to construct the spreadsheet and 22 (3 working days) to determine ICT values for all yellow bases. Based on these results, Company A decided to extend the smart framework application to all production lines. To support Company A in this change, we organized training sessions to teach personnel how to use the smart framework.
6. Discussion
Regarding the specific case study proposed, the results of this work demonstrate that the A3 approach effectively identified and addressed data acquisition problems in Company A’s OEE software. These data acquisition errors were hindering accurate OEE calculations, undermining both performance analyses and the development of necessary improvement interventions in Company A. Conversely, applying the A3 approach played a significant role in diagnosing and removing inefficiencies, enabling the OEE software to receive reliable and automated input data, which facilitated operational excellence and sustainable development for Company A.
However, the results of this work are not confined to the proposed case study. While other literature studies – such as Pereira et al. (2019) – have shown similar improvements in OEE performance through the application of Lean philosophy, the findings of this paper extend beyond the immediate impacts on Company A, paving the way for broader research implications. Specifically, for the first time in the literature (cf. Section 2), this paper shows that the A3 approach is not only effective in enhancing manufacturing or service organizations in terms of plant, human, sustainability and supply chain performance. Rather, A3 can be substantiated as a winning approach for improving data input, retrieval and utilization in digitalized companies. The A3 approach, indeed, proved to identify wastes and sources of errors in data gathering and management, thereby avoiding unnecessary steps in data collection and preventing wrong data-driven calculations.
From this perspective, the contributions of this research add to the extant literature by building on top of the work by Chivukula and Pattanaik (2023). The latter authors recently prioritized groups of Lean tools (e.g. value stream mapping, Kanban, Heijunka, etc.) to support digital transitions in industrial contexts, evaluating their compatibility with I4.0 technologies. Our paper proposes an additional Lean approach (i.e. A3), which has been overlooked by Chivukula and Pattanaik (2023). The A3 approach appears generally applicable in any context, easy to implement and effective to support companies in tackling the digital transformation challenge. More in general, the contributions of this paper corroborate the findings of Frecassetti et al. (2024), who recently explored how Lean, as a concept, can overcome barriers and support the I4.0 transition in companies undergoing digitalization. In particular, this work not only reinforces Frecassetti et al.’s claims but also introduces a practical tool (i.e. A3) that can facilitate successful digital transformation in companies.
Last but most importantly, the contributions of this study reinforce those of Torri et al. (2021), affirming the necessity of investing in Lean approaches, particularly A3, to enhance procedures and processes in the IT departments of digitalized companies. Indeed, this paper shows that the momentaneous implementation of digital solutions (e.g. OEE software) in companies is insufficient for achieving the I4.0 transition, as well as operational excellence and sustainable development. Rather, continuous support is essential during and after the implementation phases to ensure consistency. Specifically, the present study shows, for the first time that, the A3 approach can serve as a valuable tool to ensure the reliability of input data for feeding technologies and software. This aspect is crucial for avoiding incorrect data-driven analyses of company performance, in turn preventing the identification of erroneous inefficiencies, and ultimately hindering a company’s ability to strive for continuous improvement.
7. Conclusion
Equipping industrial plants with digital technologies and software allows for collecting data from the field, discovering hidden patterns, and facilitating data-driven improvement decisions. While digital technologies and software pave the way for automated continuous improvement, the sole installation of technologies and software falls short of attaining operational excellence and sustainable development. Another crucial factor involves supplying digital solutions with reliable data. To examine data retrieval processes, pinpoint error sources and implement corrections, companies should embrace structured problem-solving approaches, among which A3 is a powerful one. However, the extant literature overlooks A3 applications in information management contexts. To fill this gap, this paper proposes a case study of A3 application in Company A.
Before the A3 application, Company A had installed a new OEE software to monitor its production performance. Trusting this software and not questioning the reliability of its input data, Company A believed to have an OEE greater than 1. However, the A3 application highlighted that, despite investments in advanced software, OEE measurements were undermined by errors in data supply. This left Company A unaware of current inefficiencies, which hindered the pursuit of operational excellence and sustainability. Besides detecting data, A3 enabled Company A to identify the problem’s root causes and countermeasures. Focusing on implementing one countermeasure, within seven months Company A was able to obtain a new procedure (herein called “smart framework”) for automatically collecting data, calculating OEE, and highlighting opportunities for improvement. Besides rectifying data collection errors, the automated nature of the smart framework saved time in evaluating company performance, laying the groundwork for real-time OEE calculations. The implementation of the smart framework was tested in one production line, revealing an OEE value of 0.8. Targeting the inefficiencies identified by the smart framework, a simulation and a DOE were conducted to test in vitro potential improvement interventions. The optimal improvement solution was determined using Minitab’s Response Optimizer. This solution was field-tested, confirming the achievement of the project’s must-have and nice-to-have targets.
This paper presents both theoretical and practical contributions. Theoretically, this case study serves as empirical evidence that the sole implementation of advanced technologies and software in industrial settings is insufficient for achieving operational excellence and sustainability. Instead, the realization of smart manufacturing environments, conducive to continuous improvement, hinges on providing digital solutions with reliable input data. Therefore, the application of structured problem-solving approaches becomes imperative to identify and rectify issues in data retrieval processes, with the present case study corroborating the efficacy of the A3 approach in this regard. On a practical level, the proposed case study illustrates how A3 has been instrumental in Company A for pinpointing unnoticed issues in data management. This facilitated an increase (+10%) in the performance of a production line. A3 also led to a reduction in OEE calculation times, paving the way for real-time analyses. Finally, an auxiliary contribution of this paper is the introduction of a novel smart framework stemming from the A3 application. This smart framework allows companies to automate input data collection and ICT values for OEE calculation.
However, this research has limitations. First, it relies on a single case study and does not extend the analysis to other companies and industrial sectors. Second, while the A3 approach applied in Company A identified numerous root causes and corresponding countermeasures, only one countermeasure was implemented for time and budget reasons. Furthermore, the effectiveness of the developed countermeasure was validated through both simulation and field testing. Nevertheless, the simulation was based on a DOE with only two replications, and the field analyses involved testing the smart framework on a single production line. To address these limitations, future research developments could be envisioned. First, the A3 approach could be tested in other production lines within Company A, and in other companies. Exploring additional case studies could help generalize the paper’s findings. Second, the implementation of other countermeasures in Company A could be explored to gauge the extent of improvement in OEE and reaffirm the A3 validity. Finally, building upon the same case study, a comparative analysis could be undertaken to contrast the A3 approach with alternative problem-solving approaches, highlighting similarities and differences.
Figures
Proposed countermeasures
Root causes to be removed | Countermeasures |
---|---|
No full knowledge of working time | Gemba walks; smart framework |
Wrong values from sensors | Sensors’ replacement |
No fostering of benefits of OEE | OEE graph |
Variability in the process | |
Wrong reference parameters and input data | Gemba walks; smart framework |
No-user friendly interface | OEE graph |
Data loss during transfer to OEE software | Smart framework |
No smart way to collect data | Smart framework |
Source: Authors’ own work
Performance measured by sensors in the loading and pre-dispersion phases
Sub-phase ID | Sub-phase name | Measured performance | Description |
---|---|---|---|
101 | Loading liquids | Tank weight | Liquid raw materials are loaded (experiencing a weight increase) |
102 | Start standby | Time | A standby is inserted to stop the automatic flow and the production |
103 | End standby | Time | A standby is inserted to stop the automatic flow and the production |
104 | Loading solids | Tank weight up | Solid raw materials are loaded (experiencing a weight increase) |
105 | Mixing | Pump velocity, energy consumed | Raw materials are mixed through a pump (no weight changes) |
106 | Transfer | Tank weight | The product is transferred to Tank 1 |
107 | Loading liquids | Tank weight, pump pressure | Cleaning solvent is loaded and recirculated through a pump |
108 | Transfer | Tank weight down | The solvent is transferred to Tank 1 (with a weight-reduction) |
Source: Authors’ own work
Example of ICT estimation for a specific base in a line
Order | Cycle time registered | Order | Cycle time registered |
---|---|---|---|
1012832655 | 00:12:00 | 1012832655 | 00:12:00 |
1012832656 | 00:11:00 | 1012832656 | 00:11:00 |
1012832657 | 00:09:00 | 1012832657 | 00:13:00 |
1012832658 | 00:10:00 | 1012832658 | 00:13:00 |
1012832659 | 00:10:00 | 1012882891 | 00:15:00 |
1012865448 | 00:11:00 | 1012882892 | 00:13:00 |
1012865449 | 00:10:00 | 1012882893 | 00:11:00 |
1012865450 | 00:07:00 | 1012882894 | 00:59:00 |
1012865471 | 00:10:00 | 1012882895 | 00:16:00 |
1012865472 | 00:13:00 | 1012882896 | 00:14:00 |
1012865473 | 00:12:00 | 1012865471 | 00:12:00 |
1012882808 | 00:10:00 | 1012865472 | 00:12:00 |
1012882810 | 00:10:00 | 1012865473 | 00:11:00 |
1012882891 | 00:14:00 | 1012882895 | 00:05:00 |
1012882892 | 00:15:00 | 1012882896 | 00:12:00 |
1012882893 | 00:11:00 | 1012882897 | 00:12:00 |
1012882894 | 00:14:00 | 1012882898 | 00:11:00 |
1012882895 | 00:10:00 | 1012882899 | 00:10:00 |
1012882896 | 00:15:00 | 1012865460 | 00:10:00 |
1012882897 | 00:11:00 | 1012865461 | 00:25:00 |
1012882893 | 00:11:00 | 1012865462 | 00:10:00 |
1012882894 | 00:11:00 | 1012865463 | 00:05:00 |
1012882895 | 00:10:00 | ||
Mean cycle time | Standard deviation In cycle times |
Minimum Cycle time registered |
ICT |
00:12:39 | 00:07:42 | 00:05:00 | 00:05:00 |
Source: Authors’ own work
DOE with two factors and two replications
RPM [round/min] |
Pump speed [m/s] |
Energy rate [W*min] |
Temperature [°C] |
Flow rate [m3/min] |
Pressure [bar] |
---|---|---|---|---|---|
540 | 45 | 270 | 47.9 | 3,340 | 1.0 |
540 | 55 | 275 | 47.9 | 4,117 | 1.1 |
560 | 45 | 260 | 47.2 | 3,350 | 1.0 |
560 | 55 | 255 | 44.5 | 5,000 | 0.8 |
540 | 45 | 269 | 49.5 | 3,600 | 1.0 |
540 | 55 | 276 | 48.5 | 4,400 | 1.0 |
560 | 45 | 260 | 49.4 | 3,700 | 1.0 |
560 | 55 | 257 | 44.5 | 5,800 | 0.9 |
Source: Authors’ own work
Appendix. A3 report
References
Akbulut, Y., Usubütün, A., Durur, F. and Kutlu, G. (2023), “Reducing turnaround time in a pathology laboratory using the lean methodology”, International Journal of Lean Six Sigma, Vol. 14 No. 3, pp. 520-533, doi: 10.1108/IJLSS-12-2021-0213.
Al Busaidi, R.Z. and Al-Busaidi, K.A. (2021), “Applying lean methodologies in knowledge creation: a case in the energy sector”, Presented at the International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management, IC3K – Proceedings, pp. 169-176.
AlKhader, W., Jayaraman, R., Salah, K., Sleptchenko, A., Antony, J. and Omar, M. (2023), “Leveraging blockchain and NFTs for Quality 4.0 implementation in digital manufacturing”, Journal of Manufacturing Technology Management, Vol. 34 No. 7, pp. 1208-1234, doi: 10.1108/JMTM-05-2023-0172.
AlKoliby, I.S.M., Abdullah, H.H. and Suki, N.M. (2023), “Linking knowledge application, digital marketing, and manufacturing SMEs’ sustainable performance: the mediating role of innovation”, Journal of the Knowledge Economy, Vol. 15 No. 2, doi: 10.1007/s13132-023-01157-4.
Apaza-Casabona, G., Bacilio-Peve, S., Leon-Chavarri, C. and Ramos-Palomino, E. (2022), “Designing a lean-based production management model to reduce non-fulfilled orders at a metalworking company”, Presented at the Proceedings - 2022 8th International Conference on Information Management, ICIM 2022, pp. 135-140, doi: 10.1109/ICIM56520.2022.00031
Battistella, C., Fornasier, A. and Pessot, E. (2023), “How can lean tools support the innovation process of SMEs?”, Journal of Manufacturing Technology Management, Vol. 34 No. 6, pp. 1004-1024, doi: 10.1108/JMTM-12-2022-0449.
Bevilacqua, M., Ciarapica, F.E. and De Sanctis, I. (2017), “Relationships between Italian companies’ operational characteristics and business growth in high and low lean performers”, Journal of Manufacturing Technology Management, Vol. 28 No. 2, pp. 250-274, doi: 10.1108/JMTM-02-2016-0024.
Cantini, A., De Carlo, F. and Tucci, M. (2020), “Application of the lean layout planning system in a leather bags manufacturing plant and proposal of an approach to engage the company’s staff in the research of the layout solution”, Presented at the Proceedings of the Summer School Francesco Turco.
Carpentari, A., Cattaneo, D., Milani, M. and Rossi, M. (2020), “Frictionless Innovative Slick® process to keep track of companies’ knowledge”, Lecture Notes in Networks and Systems, Vol. 122, pp. 385-394, doi: 10.1007/978-3-030-41429-0_38.
Chivukula, K. and Pattanaik, L.N. (2023), “Prioritizing lean tools for integration with Industry 4.0 technologies using a hybrid fuzzy Moora approach”, International Journal of Industrial Engineering : Theory Applications and Practice, Vol. 30, pp. 1133-1149, doi: 10.23055/ijietap.2023.30.5.8883.
Eisenhardt, K.M. and Graebner, M.E. (2007), “Theory building from cases: opportunities and challenges”, Academy of Management Journal, Vol. 50 No. 1, pp. 25-32, doi: 10.5465/amj.2007.24160888.
Ferraro, S., Cantini, A., Leoni, L. and De Carlo, F. (2023), “Sustainable logistics 4.0: a study on selecting the best technology for internal material handling”, Sustainability, Vol. 15 No. 9, p. 7067, doi: 10.3390/su15097067.
Ferrazzi, M., Frecassetti, S. and Staudacher, A.P. (2023), “Influence of the lean approach on corporate environmental sustainability: a case study”, IFIP Advances in Information and Communication Technology 668 IFIP, pp. 285-294, doi: 10.1007/978-3-031-25741-4_24.
Fiorello, M., Gladysz, B., Corti, D., Wybraniak-Kujawa, M., Ejsmont, K. and Sorlini, M. (2023), “Towards a smart lean green production paradigm to improve operational performance”, Journal of Cleaner Production, Vol. 413, doi: 10.1016/j.jclepro.2023.137418.
Frecassetti, S., Ferrazzi, M. and Portioli-Staudacher, A. (2023), “Lean as a facilitator for AGVs implementation: a case study”, IFIP Advances in Information and Communication Technology 668 IFIP, pp. 95-107, doi: 10.1007/978-3-031-25741-4_10.
Frecassetti, S., Rossini, M. and Portioli-Staudacher, A. (2024), “Unleashing Industry 4.0: Leveraging lean practices to overcome implementation barriers”, IEEE Transactions on Engineering Management, Vol. 71 No. 20, doi: 10.1109/TEM.2024.3396448.
Fuenzalida, C., Fischer, B., Arroyo, P. and Salvatierra, J.L. (2016), “Evaluating environmental impacts of construction operation before and after the implementation of lean tools”, Presented at the, IGLC 2016 – 24th Annual Conference of the International Group for Lean Construction, pp. 3-12.
Gijo, E.V., Palod, R. and Antony, J. (2018), “Lean six sigma approach in an Indian auto ancillary conglomerate: a case study”, Production Planning and Control, Vol. 29 No. 9, pp. 761-772, doi: 10.1080/09537287.2018.1469801.
Gnanaguru, R., Puvaneswari, K., Mallick, J., Jegadheesan, C., Sivakumar, V.M. and Devadasan, S.R. (2011), “Toyota’s A3 reports for improving 6-S activities: an aeronautical industry case study”, International Journal of Services and Operations Management, Vol. 10 No. 2, pp. 239-254, doi: 10.1504/IJSOM.2011.042519.
Henriquez, R., Muñoz-Villamizar, A. and Santos, J. (2023), “Key factors in operational excellence for Industry 4.0: an empirical study and maturity model in emerging countries”, Journal of Manufacturing Technology Management, Vol. 34 No. 5, pp. 771-792, doi: 10.1108/JMTM-09-2022-0330.
Kanyinda, K., Lazarus, I.J. and Olanrewaju, O.A. (2020), “A3 problem solving: a case of assembly line downtime”, Presented at the Proceedings of the International Conference on Industrial Engineering and Operations Management, pp. 284-295.
Karlsson, S., Oghazi, P., Hellstrom, D., Patel, P.C., Papadopoulou, C. and Hjort, K. (2023), “Retail returns management strategy: an alignment perspective”, Journal of Innovation and Knowledge, Vol. 8 No. 4, p. 100420, doi: 10.1016/j.jik.2023.100420.
Kassem, B., Rossini, M., Costa, F. and Portioli-Staudacher, A. (2023), “Lean monitoring: action research in manufacturing”, International Journal of Lean Six Sigma, Vol. 14 No. 6, pp. 1280-1296, doi: 10.1108/IJLSS-06-2022-0124.
Khodambashi, S. (2015), “Alignment of an intra-operating management process to a health information system: a lean analysis approach”, Personal and Ubiquitous Computing, Vol. 19 No. 3-4, pp. 689-698, doi: 10.1007/s00779-015-0853-0.
Koreček, D., Solfronk, P. and Sobotka, J. (2020), “Numerical simulation as a tool to predict sheet metal forming process of TRIP steel HCT690”, Manufacturing Technology, Vol. 20 No. 5, pp. 625-631, doi: 10.21062/MFT.2020.103.
Krogstie, L., Ulonska, S., Welo, T. and Andersen, B. (2014), “On knowledge-based development: how documentation practice represents a strategy for closing tolerance engineering loops”, Procedia CIRP, Vol. 21, pp. 318-323, doi: 10.1016/j.procir.2014.03.156.
Li, N., Laux, C.M. and Antony, J. (2019), “How to use lean six sigma methodology to improve service process in higher education: a case study”, International Journal of Lean Six Sigma, Vol. 10 No. 4, pp. 883-908, doi: 10.1108/IJLSS-11-2018-0133.
Low, M.Y.H., Chan, D.Y.K., Lee, K.K., Soh, D.C.L., Lin, D.W. and Wee, L.B. (2022), “Online team-based learning for productivity management module”, Presented, at the Proceedings - 2022 IEEE International Conference on Teaching, Assessment and Learning for Engineering, TALE 2022, pp. 436-440, doi: 10.1109/TALE54877.2022.00077
Marques, P.A., Carvalho, A.M. and Santos, J.O. (2022), “Improving operational and sustainability performance in a retail fresh food market using lean: a Portuguese case study”, Sustainability, Vol. 14 No. 1, p. 403, doi: 10.3390/su14010403.
Mishrif, A. and Khan, A. (2023), “Technology adoption as survival strategy for small and medium enterprises during COVID-19”, Journal of Innovation and Entrepreneurship, Vol. 12 No. 1, doi: 10.1186/s13731-023-00317-9.
Muhuri, P.K., Shukla, A.K. and Abraham, A. (2019), “Industry 4.0: a bibliometric analysis and detailed overview”, Engineering Applications of Artificial Intelligence, Vol. 78, pp. 218-235, doi: 10.1016/j.engappai.2018.11.007.
Naik, T., Duroseau, Y., Zehtabchi, S., Rinnert, S., Payne, R., McKenzie, M. and Legome, E. (2012), “A structured approach to transforming a large public hospital emergency department via lean methodologies”, Journal for Healthcare Quality, Vol. 34 No. 2, pp. 86-97, doi: 10.1111/j.1945-1474.2011.00181.x.
Nguyen, D.M. (2015), “A new application model of lean management in small and medium sized enterprises”, International Journal of Simulation Modelling, Vol. 14 No. 2, pp. 289-298, doi: 10.2507/IJSIMM14(2)9.304.
Pereira, J., Silva, F.J.G., Bastos, J.A., Ferreira, L.P. and Matias, J.C.O. (2019), “Application of the A3 methodology for the improvement of an assembly line”, Procedia Manufacturing, Vol. 38, pp. 745-754, doi: 10.1016/j.promfg.2020.01.101.
Powell, D.J. (2024), “Artificial intelligence in lean manufacturing: digitalization with a human touch?”, International Journal of Lean Six Sigma, Vol. 15 No. 3, pp. 719-729, doi: 10.1108/IJLSS-05-2024-256.
Roche, K.E. and Baumgartner, R.J. (2024), “Development of a strategy deployment framework combining corporate sustainability and operational excellence”, Corporate Social Responsibility and Environmental Management, Vol. 31 No. 3, pp. 2159-2174, doi: 10.1002/csr.2683.
Rojas, K.K.P., Quiroz-Flores, J.C. and Nallusamy, S. (2023), “Minimization of product distribution delays through an integration model of lean manufacturing tools and A3 report-case study”, International Journal of Mechanical Engineering, Vol. 10 No. 9, pp. 31-43, doi: 10.14445/23488360/IJME-V10I9P103.
Rossini, M., Cifone, F.D., Kassem, B., Costa, F. and Portioli-Staudacher, A. (2021), “Being lean: how to shape digital transformation in the manufacturing sector”, Journal of Manufacturing Technology Management, Vol. 32 No. 9, pp. 239-259, doi: 10.1108/JMTM-12-2020-0467.
Santhiapillai, F.P. and Ratnayake, R.M.C. (2021), “A lean approach for multi-criteria decision-making in public services’ strategy deployment”, IFIP Advances in Information and Communication Technology 630 IFIP, pp. 656-664, doi: 10.1007/978-3-030-85874-2_71.
Santhiapillai, F.P. and Ratnayake, R.M.C. (2023), “Lean thinking and strategy deployment: adapting Hoshin Kanri and A3-based project prioritization in police services”, The TQM Journal Ahead-of-Print, doi: 10.1108/TQM-05-2022-0177.
Santos Filho, G.M. and Simão, L.E. (2022), “A3 methodology: going beyond process improvement”, Revista de Gestão, Vol. 30 No. 2, pp. 147-161, doi: 10.1108/REGE-03-2021-0047.
Shook, J. (2008), Managing to Learn: Using the A3 Management Process to Solve Problems, Gain Agreement, Mentor and Lead, Lean Enterprise Institute.
Singh, V. and Muller, G. (2013), “Knowledge capture, cross boundary communication and early validation with dynamic A3 architectures”, Presented at the 23rd Annual International Symposium of the International Council on Systems Engineering, INCOSE 2013, pp. 420-432.
Sobek, D.K. and Jimmerson, C. (2004), “A3 reports: tool for process improvement”, IIE Annual Conference. Proceedings. Institute of Industrial and Systems Engineers (IISE), p. 1.
Sobek, D.K. and Jimmerson, C. (2006), “A3 reports: tool for organizational transformation”, IIE Annual Conference. Proceedings 1–6.
Sunder, M.V., Mahalingam, S. and Krishna M, S.N. (2020), “Improving patients’ satisfaction in a mobile hospital using lean six sigma – a design-thinking intervention”, Production Planning and Control, Vol. 31 No. 6, pp. 512-526, doi: 10.1080/09537287.2019.1654628.
Terelak-Tymczyna, A. and Niesterowicz, B. (2024), “Lean manufacturing in digital transformation of manufacture”, Lecture Notes in Mechanical Engineering, pp. 74-83, doi: 10.1007/978-3-031-56444-4_6.
Torri, M., Kundu, K., Frecassetti, S. and Rossini, M. (2021), “Implementation of lean in IT SME company: an Italian case”, International Journal of Lean Six Sigma, Vol. 12 No. 5, pp. 944-972, doi: 10.1108/IJLSS-05-2020-0067.
Tsukada, O., Ibusuki, U., Kuchii, S. and de Almeida, A.T.S.B. (2024), “Is lean manufacturing maturity a prerequisite for Industry 4.0? Survey of SMEs in Japan and Brazil”, International Journal of Lean Six Sigma, Vol. 15 No. 5, doi: 10.1108/IJLSS-08-2023-0149.
Virmani, N. and Ravindra Salve, U. (2023), “Significance of human factors and ergonomics (HFE): mediating its role between Industry 4.0 implementation and operational excellence”, IEEE Transactions on Engineering Management, Vol. 70 No. 11, pp. 3976-3989, doi: 10.1109/TEM.2021.3091398.
Voss, C. (2008), Case Research in Operations Management, Researching Operations Management. Routledge.
Wang, J., Tsiligiris, V. and Hartley, R. (2019), “Perceptions of the benefits of the A3 planner in facilitating project-based learning in accounting education”, Accounting Education, Vol. 28 No. 3, pp. 221-238, doi: 10.1080/09639284.2018.1516151.
Yin, R.K. (2018), “Case study research: design and methods”, 6th ed., Applied Social Research Methods Series, SAGE, Thousand Oaks, CA.
Yin, Y., Stecke, K.E. and Li, D. (2018), “The evolution of production systems from Industry 2.0 through Industry 4.0”, International Journal of Production Research, Vol. 56 Nos 1/2, pp. 848-861, doi: 10.1080/00207543.2017.1403664.
Yusof, M.M., Khodambashi, S. and Mokhtar, A.M. (2012), “Evaluation of the clinical process in a critical care information system using the lean method: a case study”, BMC Medical Informatics and Decision Making, Vol. 12 No. 1, p. 150, doi: 10.1186/1472-6947-12-150.
Zuñiga, R.M., Carvajal, J.E., Aldama, M. and Lopez, J.M.J. (2023), “Improving complex information flows at a transactional process using a digital-enhanced lean methodology: a case study for a Mexican-based truck-making company”, Presente7d at the IISE Annual Conference and Expo 2023, doi: 10.21872/2023IISE_2364.
Acknowledgements
This study was carried out within the MICS (Made in Italy – Circular and Sustainable) Extended Partnership and received funding from the European Union Next-GenerationEU [PIANO NAZIONALE DI RIPRESA E RESILIENZA (PNRR) – MISSIONE 4 COMPONENTE 2, INVESTIMENTO 1.3 – D.D. 1551.11–10-2022, PE00000004]. This manuscript reflects only the authors’ views and opinions, neither the European Union nor the European Commission can be considered responsible for them.
This research is also in collaboration with the HumanTech Project, which is financed by the Italian Ministry of University and Research (MUR) for the 2023–2027 period as part of the ministerial initiative “Departments of Excellence” (L. 232/2016).
Corresponding author
About the authors
Alessandra Cantini is an Assistant Professor at the Politecnico di Milano (POLIMI). She is the author and co-author of more than 20 publications in international conferences and journals. Her research areas of interest include warehouse management, supply chain management, spare parts management, additive manufacturing, operational excellence and lean manufacturing.
Federica Costa is an Assistant Professor at the Politecnico di Milano (POLIMI). She is the author and co-author of more than 35 publications in international conferences and journals. Her research areas of interest include workload control, lean management, sustainability, simulation, operational excellence and continuous improvement. Federica Costa is the corresponding author and can be contacted at: federica.costa@polimi.it
Alberto Portioli-Staudacher is a Full Professor at the Politecnico di Milano. He is also the leader of the Lean Excellence Center at (POLIMI). He is the author and co-author of more than 90 publications in international conferences and journals. His research areas of interest include lean management, operational excellence, operations management, industrial management and workload control.