Barriers to Implementing Lean Six Sigma in the Chemical Process Industry: The Case of Brazil

Abstract

Lean Six Sigma (LSS) is recognized as a powerful process improvement methodology for enhancing operational efficiency and long-term sustainability. This study examined the barriers hindering LSS implementation in the chemical process industry within an emerging economy, Brazil. It developed a structured methodology for designing mitigation measures to overcome those barriers. First, 26 barriers from the literature were ranked by LSS experts through a Delphi survey to select the top 15. Then, a combined interpretive structural modeling (ISM) approach and impact matrix cross-reference multiplication applied to a classification (MICMAC) analysis approach was used, supported by a focus group, to determine the hierarchical relationships among the barriers and their driving power and dependence. Finally, a second focus group defined adequate mitigation measures. The top four main barriers are the lack of time, insufficient systemic understanding of lean principles, misalignment between LSS and corporate strategies, and inadequate top management commitment. Additionally, 10 mitigation measures are proposed. This study contributes to LSS implementation in the chemical process industry in Brazil, thus enhancing industry sustainability by improving operational efficiency, curbing waste, reducing transportation-related emissions through a decreasing reliance on imported chemical products, and contributing to economic growth and job creation within the industry.

1. Introduction

The chemical process industry (CPI) is vital in producing daily-use products worldwide and has a long-standing history. The CPI comprises a range of industrial processes, including mixing, separating, forming, and chemical reactions, which primarily produce non-discrete materials such as liquids, pulps, gases, powders, and slurries [1]. The current market globalization, imposing fierce competition, has challenged CPI companies to think of new solutions and orientations to enhance their operational performance. These companies aim to reduce production, distribution, and marketing costs while satisfying clients’ requirements [2].

The CPI in Brazil, representing around 2.4% of Brazilian GDP, urgently needs to adopt new methodologies and technologies to improve its processes, add more value to its products, and reduce production costs. This would improve its performance and increase its competitiveness in the face of external competition [3]. In 2023, the CPI in Brazil was expected to have a trade deficit, exporting around USD 12.2 billion and importing USD 52 billion [4]. LSS is widely recognized as a fundamental tool to sustain competitiveness, achieve long-lasting results, and improve production [5,6]. Thus, promoting Lean Six Sigma (LSS) implementation in the Brazilian CPI can help reduce its current external trade deficit, improving the Brazilian balance of payments (BOP). Moreover, due to cultural changes and social pressures, the impacts of economic activities are increasingly under increased scrutiny. As a result, organizations are compelled to enhance their performance by improving their environmental and societal impacts [7].

Despite its advantages, organizations still have difficulties implementing and maintaining LSS [2]. In the literature, several studies address the issue of adopting LSS or related methodologies [1,8,9,10,11]. However, there is still a lack of research on adopting LSS in the CPI [2,12]. LSS’s roots are in the discrete manufacturing industry, which aggravates the challenges of implementing it in the CPI. The CPI has distinctive production processes characterized by continuous flows and batch steps, which can be influenced by factors like temperature, process variables, and chemical or physical characteristics [2]. Scheller et al. [13], when investigating the case study of a Brazilian industrial company, concluded that barriers to LSS implementation in developing countries should be further studied.

This study aimed to bridge the gap in knowledge regarding LSS implementation in the CPI by investigating the related barriers in an emerging economy, Brazil, and suggesting an innovative methodology for designing effective and practical mitigation measures (MMs). The research questions were the following: What are the main barriers to LSSI in the CPI? What are the hierarchical relationships between those barriers to LSS? What are the driving power and dependence of those barriers?

A combined interpretive structural modeling (ISM) and impact matrix cross-reference multiplication applied to a classification (MICMAC) analysis approach, supported by Delphi surveys and focus group discussions, was adopted to address the above questions. This approach can explore the relationships between barriers and their driving power and dependence and has been proven in similar studies [14].

This study expands the literature in two regards, first, by proposing a structured methodology to identify the main barriers to LSS implementation (LSSI) in the CPI, map the hierarchical relationships between them, and ascertain their driving power and dependence. This approach facilitates the development of measures to address an interconnected system of barriers, rather than addressing them in isolation. Second, it provides a novel perspective on a developing country. This study’s findings have practical and professional implications in allowing the CPI and governments to define a roadmap of actions to mitigate the main barriers they face, thus improving the conditions for successful LSSI.

This paper is organized as follows: Section 2 comprises the literature review; Section 3 covers the research methodology; Section 4 presents the results; Section 5 discusses the results and presents the MMs; and, finally, Section 6 presents the main results, main contributions and conclusions, limitations, and future research.

2. Literature Review

This section encompasses the literature review on LSS, the CPI, and barriers to the adoption of LSS in the CPI and ends with the justification for the present study.

2.1. Lean Six Sigma

Lean approaches, grounded in the Toyota Production System (TPS) [15], focus on removing non-value-adding activities, or waste, from the process through a continuous improvement and flow-oriented approach. These are centered on identifying value, identifying value streams, promoting flow, adopting pull systems, and seeking perfection to reduce the flow time of processes. They assume that waste removal improves organizational performance, and the premise is that minor improvements are preferable to systems analysis [16,17]. Meanwhile, Six Sigma, developed at Motorola in the 1980s, aims to reduce variability and defects through a problem-focused definition and statistical control of outputs, based on defining, measuring, analyzing, improving, and controlling (DMAIC) processes [18,19]. This approach requires statistical tools and robust analytical techniques such as statistical process control, quality function deployment, failure mode and effect analysis (FMEA), analysis of variance, design of experiments, and the Kano model. LSS arose from the idea of combining these two distinct approaches that complement each other [5].

The term LSS first appeared around 2000. LSS offers better results by improving organizations’ capacity to reduce production costs and maximize value by enhancing quality [16]. As the most advanced approach to increasing a company’s competitiveness, LSS has emerged as the leading methodology chosen by industries [18]. The adoption of LSS has led to many transformative improvements, shaping the landscape of industries and work culture. These improvements encompass new organizational structures and operational approaches, greater employee engagement and knowledge, greater transparency, the introduction of diverse metrics to evaluate company and employee performance, a reduction in siloed functions in favor of more transactional functions, cost savings, fewer production stops, fewer defects, and augmented product value. These improvements lead to enhanced quality and production efficiency [5]. LSS exerts a profound influence, extending its impact across a broad spectrum of internal and external processes that are influenced by all facets of production.

2.2. Chemical Process Industry

Companies in the CPI, a subcategory of the process industry sector, transform raw materials into new products through chemical and physical reactions, a process that cannot be reversed [1]. As its importance has grown, the CPI has continually adapted its production methods in response to changing client needs, increasing competition among companies, fluctuations in demand, and the emergence of new technologies; this presents constant challenges for this sector and necessitates a constant search for new tools, adaptations, and improvements.

Production is the primary purpose of any process plant, not only chemical plants, and this factor ultimately determines whether the plant thrives or falters. Therefore, one of the foremost objectives of any chemical process plant is to attain the utmost efficiency and cost-effectiveness, minimizing interruptions or machinery issues while ensuring the highest possible product quality [1]; this has become an increasingly important point of research focus in recent decades due to globalization. However, the production process is inherently intricate, involving numerous steps, many technologies, and various stakeholders, which presents a perpetual challenge in the pursuit of uninterrupted and seamless production [12].

In this context, the CPI has demonstrated the need for new production methodologies, especially agile ones, as well as tools and cultural changes [2]. LSS has been proposed as a method for improving manufacturing and service processes [20], even though some authors argue that its application to the CPI is not straightforward [12].

2.3. Barriers to LSS Implementation

While the Lean methodology has numerous derivatives, such as production, manufacturing, startup, and others, each with distinct features, particularly in implementation processes, they all share the same underlying Lean philosophy, objectives, and characteristics; the same applies to the Six Sigma methodology. Furthermore, there are often many common barriers encountered in implementing Lean, Six Sigma, and LSS. Based on a literature review, Table 1 presents a sample of studies from the last decade considered more relevant to research on the barriers to the implementation of LSS and related methodologies.

Several studies address the barriers to implementing methodologies related to Lean [1,14,21,22,23,24,25,26,27,28]. Only a few studies address the barriers to implementing LSSI [9,10,20,29,30]. Twelve studies were developed in India, one in the U.S., one in Norway, one in Brazil, one in South Africa, and one in Bangladesh. Methodologically, a systematic literature review was used in six studies, a questionnaire survey in five (sometimes combined with other methodologies), an ISM and MICMAC analysis in six, a structural equation modeling (SEM) in four, an exploratory factor analysis (EFA) in one, and Liker’s 4P model for Lean philosophy in one. Finally, most of the studies were carried out in the context of the manufacturing industry, with the exception of two studies, which concerned the process industry [1,26], of which the CPI is a part.

Table 1. Studies from the last decade on barriers to Lean and LSS implementation.

Authors	Main Findings
Jadhav, [21]	Identified 24 barriers to Lean implementation across sectors and geographies. The top barriers were a lack of investment resources, senior management involvement, and workers’ resistance to Lean. Emphasized understanding the correlations between barriers and collective mitigation.
Panwar et al. [1]	Found Lean to be rarely adopted in Indian process industries, especially in the chemical sector. The main barriers included small production batches, hiring Lean experts, communication, and employee training. Lean was valued for reducing waste and increasing quality when adopted.
Abolhassani et al. [27]	Highlighted differences between Lean practitioners and nonpractitioners in the U.S. Practitioners identified a lack of technical knowledge and employee resistance as critical issues. On the other hand, nonpractitioners viewed top management commitment and Lean’s suitability for continuous processes as the main barriers to implementation.
Lodgaard et al. [22]	A two-year case study found that top managers, middle managers, and workers in a Norwegian company faced different barriers to Lean implementation. The barriers were categorized under management, organization for Lean, Lean tools, and knowledge. Success was limited due to these multilevel challenges.
Panwar et al. [26]	Found that lack of familiarity, expertise, education, training, and management support were major reasons for not implementing Lean in India’s process industry. Unique process characteristics and cultural barriers were considered less significant factors.
Muganyi et al. [12]	Reviewed the tools for LSSI in the CPI in South Africa. Concluded that LSSI effectively improves performance, including financial and quality gains. The strategic positioning of LSSI is critical for innovation and success in the CPI.
Kaswan and Rathi [9]	Analyzed enablers for Green LSS implementation. Identified data assimilation, Lean Green matrices, and organizational readiness as critical enablers for successful GLSS.
Rathi et al. [29]	Prioritized the barriers to LSSI in Indian micro, small, and medium enterprises (MSMEs). Management-related barriers were the most significant, followed by organization-related ones, highlighting the importance of leadership and organizational support for successful LSSI implementation.
Yadav et al. [23]	Explored Lean implementation barriers in SMEs in India. The main barriers were a lack of management commitment, leadership, and resources, poor communication, and insufficient knowledge of Lean’s benefits. Findings were consistent with larger organizations.
Ali et al. [20]	Developed a framework applied to an apparel manufacturing company. Found that lack of training, insufficient investments, and poor information sharing were the main barriers to LSSI implementation in Bangladesh. Addressing these issues was critical for success.
Chaple et al. [14]	Prioritized the barriers to Lean implementation in India. The most critical were insufficient management time, supervisory skills, and senior management skills. Cost and funding-related barriers were considered less significant.
Ramadas and Satish [30]	Identified the barriers to Lean in SMEs in India, including inadequate training programs, low-quality materials from suppliers, and poor communication between supervisors and workers.
Abu et al. [24]	The key barriers to Lean implementation in manufacturing industries in India were found to be a lack of understanding of Lean, insufficient resources, and employee resistance. Synergetic effects were identified between cultural, human attitudinal, and resource-related issues.
Swarnakar et al. [10]	Developed a hierarchical model to assess critical success factors (CSFs) for sustainable LSS in hospitals. Environmental and social responsibility, senior management leadership, and efficient resource usage were key factors.
Leite et al. [11]	Categorized Lean barriers into six groups: behavioral/cultural, organizational strategy, leadership, technical limitations, process-based issues, and resource constraints. Proposed eight measures to mitigate barriers.
Puram et al. [25]	Provided a systemic understanding of Lean barriers across sectors, including lack of resources, expertise, training, Lean principles knowledge, and co-ordination issues.
Patel and Patel [31]	Investigated LSSI adoption in Indian SMEs. Found that consultants are crucial for awareness and skills development. Top management commitment is vital, while resistance to change is a significant challenge.
Mano et al. [28]	Identified seven critical factors for Lean Construction in Brazil: team co-operation, performance measurement, decision-making involvement, leadership openness, and information flow.
Kumar et al. [32]	Identified and prioritized the barriers to LSSI in Indian manufacturing industries in the context of Industry 4.0 (I4.0). The main barriers include a lack of leadership, insufficient advisory and monitoring support, unclear understanding of the economic benefits, and inadequate integration of LSS with Industry 4.0 initiatives.
Abu-Salim et al. [33]	Investigated the barriers to LSSI in the service industry in India and suggested that the top barriers are lack of top management commitment, lack of customer focus, resistance to change management, and lack of alignment between the LSS and organizational strategy.
Mohan, Kaswan, and Rathi [34]	Examined the implementation of green LSS in Indian MSMEs and found that the primary barriers include insufficient top leadership support, a lack of adequate employees to carry out new strategies, and the absence of a technological plan for LSSI.

Table 1. Studies from the last decade on barriers to Lean and LSS implementation.

Authors	Main Findings
Jadhav, [21]	Identified 24 barriers to Lean implementation across sectors and geographies. The top barriers were a lack of investment resources, senior management involvement, and workers’ resistance to Lean. Emphasized understanding the correlations between barriers and collective mitigation.
Panwar et al. [1]	Found Lean to be rarely adopted in Indian process industries, especially in the chemical sector. The main barriers included small production batches, hiring Lean experts, communication, and employee training. Lean was valued for reducing waste and increasing quality when adopted.
Abolhassani et al. [27]	Highlighted differences between Lean practitioners and nonpractitioners in the U.S. Practitioners identified a lack of technical knowledge and employee resistance as critical issues. On the other hand, nonpractitioners viewed top management commitment and Lean’s suitability for continuous processes as the main barriers to implementation.
Lodgaard et al. [22]	A two-year case study found that top managers, middle managers, and workers in a Norwegian company faced different barriers to Lean implementation. The barriers were categorized under management, organization for Lean, Lean tools, and knowledge. Success was limited due to these multilevel challenges.
Panwar et al. [26]	Found that lack of familiarity, expertise, education, training, and management support were major reasons for not implementing Lean in India’s process industry. Unique process characteristics and cultural barriers were considered less significant factors.
Muganyi et al. [12]	Reviewed the tools for LSSI in the CPI in South Africa. Concluded that LSSI effectively improves performance, including financial and quality gains. The strategic positioning of LSSI is critical for innovation and success in the CPI.
Kaswan and Rathi [9]	Analyzed enablers for Green LSS implementation. Identified data assimilation, Lean Green matrices, and organizational readiness as critical enablers for successful GLSS.
Rathi et al. [29]	Prioritized the barriers to LSSI in Indian micro, small, and medium enterprises (MSMEs). Management-related barriers were the most significant, followed by organization-related ones, highlighting the importance of leadership and organizational support for successful LSSI implementation.
Yadav et al. [23]	Explored Lean implementation barriers in SMEs in India. The main barriers were a lack of management commitment, leadership, and resources, poor communication, and insufficient knowledge of Lean’s benefits. Findings were consistent with larger organizations.
Ali et al. [20]	Developed a framework applied to an apparel manufacturing company. Found that lack of training, insufficient investments, and poor information sharing were the main barriers to LSSI implementation in Bangladesh. Addressing these issues was critical for success.
Chaple et al. [14]	Prioritized the barriers to Lean implementation in India. The most critical were insufficient management time, supervisory skills, and senior management skills. Cost and funding-related barriers were considered less significant.
Ramadas and Satish [30]	Identified the barriers to Lean in SMEs in India, including inadequate training programs, low-quality materials from suppliers, and poor communication between supervisors and workers.
Abu et al. [24]	The key barriers to Lean implementation in manufacturing industries in India were found to be a lack of understanding of Lean, insufficient resources, and employee resistance. Synergetic effects were identified between cultural, human attitudinal, and resource-related issues.
Swarnakar et al. [10]	Developed a hierarchical model to assess critical success factors (CSFs) for sustainable LSS in hospitals. Environmental and social responsibility, senior management leadership, and efficient resource usage were key factors.
Leite et al. [11]	Categorized Lean barriers into six groups: behavioral/cultural, organizational strategy, leadership, technical limitations, process-based issues, and resource constraints. Proposed eight measures to mitigate barriers.
Puram et al. [25]	Provided a systemic understanding of Lean barriers across sectors, including lack of resources, expertise, training, Lean principles knowledge, and co-ordination issues.
Patel and Patel [31]	Investigated LSSI adoption in Indian SMEs. Found that consultants are crucial for awareness and skills development. Top management commitment is vital, while resistance to change is a significant challenge.
Mano et al. [28]	Identified seven critical factors for Lean Construction in Brazil: team co-operation, performance measurement, decision-making involvement, leadership openness, and information flow.
Kumar et al. [32]	Identified and prioritized the barriers to LSSI in Indian manufacturing industries in the context of Industry 4.0 (I4.0). The main barriers include a lack of leadership, insufficient advisory and monitoring support, unclear understanding of the economic benefits, and inadequate integration of LSS with Industry 4.0 initiatives.
Abu-Salim et al. [33]	Investigated the barriers to LSSI in the service industry in India and suggested that the top barriers are lack of top management commitment, lack of customer focus, resistance to change management, and lack of alignment between the LSS and organizational strategy.
Mohan, Kaswan, and Rathi [34]	Examined the implementation of green LSS in Indian MSMEs and found that the primary barriers include insufficient top leadership support, a lack of adequate employees to carry out new strategies, and the absence of a technological plan for LSSI.

2.4. Justification of the Research

The noteworthy success of LSS has led to its widespread adoption across various sectors, resulting in specialized adaptations for distinct areas of application [5,11]. Furthermore, LSS assists organizations in addressing increasing cultural and social pressures regarding sustainable management practices, particularly within the CPI [2]. LSS implementation aligns with the pillars of sustainability [12,35,36].

Nevertheless, despite its popularity, a significant adoption gap persists, particularly in sectors such as process industries [37], particularly in the CPI [2,18]. In addition, Scheller et al. [13] suggested that barriers to LSSI in developing countries should be further studied. Typically, studies on the barriers to LSSI (or related methodologies) rely on the identification and ranking of the barriers (questionnaire survey), considering the relationships between them (ISM and MICMAC analysis), or using multivariate statistical analysis techniques to study the structural relationships between them (SEM and EFA). However, they rarely present MMs for the barriers. When they do, the measures often focus on directly addressing the high-ranked or underlying barriers. Moreover, these measures often fail to address the remaining barriers comprehensively, as they are considered in isolation rather than within the context of the entire barrier system.

This study analyzes the barriers to LSSI in the CPI and develops practical and effective MMs based on the hierarchical relationships, driving power, and dependence among these barriers; to the authors’ knowledge, this has never been undertaken before, particularly in a developing country. Building on the gaps identified in prior studies, this research adopts a structured methodology to explore these barriers and propose solutions tailored to the Brazilian CPI context; it proposes a practical methodology that offers an innovative approach centered on an ISM-MICMAC analysis to determine the main barriers to LSSI and develop measures to mitigate them.

3. Methods and Materials

Social science research methods are most appropriate for research that comprises human behavior or actions present in all phases of innovation implementation [38], as is the case for LSSI in the CPI. Therefore, a methodological approach that includes methods that originate in the social sciences, such as ISM [39] and MICMAC analysis [40], supported by the Delphi survey [41] and focus group discussion (FGD) methods [42], is valuable for studying the barriers to LSSI. Such an approach permits an understanding of the multifaceted relationships and hierarchy between barriers and reveals how they influence and are influenced by each other, allowing the conceptualization of effective and practical MMs. Similar studies’ methodological approaches, which are centered on ISM and MICMAC analysis, are common [10,43].

3.1. ISM and MICMAC Analysis

ISM is a computer-supported learning process that is utilized to conduct systemic research on the relationships among a set of variables (barriers in the present work) related to a topic of a particularly complex system [39]. ISM decodes vague mental models into visible, well-defined systems and increases the understanding of the system’s variables by determining their hierarchy and relationships [44]; moreover, it considers the highest hierarchical variables as the root variables that primarily control the system’s behavior. Grounded in expert knowledge and practical experience, ISM clarifies the relationships and hierarchies among variables, helping researchers understand the level and direction of these multilayered relationships within complex systems [39]. Furthermore, as ISM integrates experts’ opinions and allows them to revise their views and adjust their appraisals, it can be effectively applied to real-life systems.

MICMAC analysis is based on the multiplication properties of matrices and is used to classify and better comprehend a set of variables from a system given their driving power and dependence [9]. The driving power represents a variable’s capacity to influence others, and the degree of dependence reflects the degree to which others influence it. A driving power-dependence diagram is created, and the variables are located in one of four clusters based on their scores: Independent (strong driving power and low dependence), Linkage (strong driving power and high dependence), Dependent (weak driving power and high dependence), and Autonomous (weak driving power and low dependence). The variables in the independent cluster influence most of the variables but are not influenced by others and are, thus, considered the more significant variables. Therefore, MICMAC analysis complements the ISM model. This approach enhances the hierarchical relationships among barriers and increases the rigor of the methodologic approach [45].

3.2. Delphi Survey

The Delphi survey is a tool that refines group opinions and judgments on complex topics; it leverages collective intelligence to improve individual opinions, especially when evidence or data are scarce. This survey allows participants to share their knowledge anonymously, compare their opinions with others, and modify them after reviewing the group’s collective opinions [41]. The Delphi method is an iterative, response-based approach for reaching a consensus among participants. The participants are allowed to reassess their opinions in consideration of the group’s feedback. There are no clear rules regarding the number of participants to be included in a Delphi survey. Needham and de Loë [46] suggest a sample of at least 10 experts.

Seventeen experts were selected for this study via the snowball sampling technique. This nonprobability sampling technique was chosen because of the challenge of finding experts knowledgeable in CPI across diverse sectors and in LSS, ideally certified in LSS or related methodologies [45]. Initially, two experienced experts were selected by the authors. The remaining experts were subsequently identified with assistance from the initially selected experts. In terms of the sector, around 29% of them were from the pharmaceuticals sector, 35% were from cosmetics, 23% were from agrochemicals (which includes fertilizers and related products), and 13% were from petrochemicals. Additionally, some experts had experience in multiple sectors. Moreover, approximately 35% of them were from Brazilian companies, whereas the remaining 35% were from international companies operating in Brazil.

In terms of experience, seven experts were in the field for more than 10 years, eight between 5 and 10 years, and two less than 5 years. All the selected experts possessed good knowledge of LSS and offered diverse opinions and views, with eight having Six Sigma certification (Green Belt or Black Belt).

3.3. Focus Group Discussion

The FGD is an exploratory research method that captures qualitative data through dynamic group direct interaction and discussion on a topic introduced by a moderator in a time-effective manner [42]. Discussions among experts about their opinions, perceptions, attitudes, and beliefs about a theory, concept, or product increase the amount of information available on a topic and generate new understandings. FGDs should include 4 to 12 experts who are carefully chosen to ensure variety and experience and to avoid bias [42]. Additionally, all the experts must have equal weight in the decision-making process; if consensus is not reached, the majority decision prevails.

One of the researchers, who was knowledgeable and experienced in LSS and the CPI, acted as the FGD moderator. His role was to stimulate discussion, ensuring a progression from general to specific issues to reduce bias and promote sincerity [42]. This study used two FGDs, FGD-1 and FGD-2, involving the same seven experts chosen from among the Delphi survey participants based on their LSS and CPI knowledge and experience.

3.4. Methodological Framework

The methodological framework adopted in this study comprises four stages (Figure 1). Expert opinions were gathered through a Delphi survey and two FGDs.

3.4.1. Stage I: Establishing the Critical Barriers to LSSI in the CPI

This stage consists of two steps. In Step 1, a literature review was conducted on the 18 most relevant publications in Table 1 to identify the common barriers to Lean, LSS, and related methodologies. Two academics knowledgeable about CPI and LSS selected and consolidated similar barriers from the literature. Then, they adapted them to LSSI and the Brazilian context, making it possible to address concerns about some issues being “lost in translation” [47]. Additionally, special attention was given to eliminating vague words and uncertainties to ensure a clear understanding for experts and enhance the validity of their opinions. The final list of 26 barriers is presented in

Table 2. Barriers to LSS implementation in the CPI, criticality score, and ranking.

No.	Barrier	GM	Rank	References
1	Lack of strategic alignment between LSS and corporate strategies for the short and long term (B3)	5.86	3	[12,14,16,22,23,25,27,29,31,35,48,49]
2	Lack of systemic understanding of LSS, the need for LSSI, and its benefits (B1)	6.21	1	[12,13,14,16,22,24,25,27,29,48,50]
3	Lack of top management commitment and supervision for LSSI (B2)	6.11	2	[12,14,22,24,25,26,27,29,35,36,50]
4	Lack of ownership of LSSI from the staff (B4)	5.79	4	[11,21,22,23,25]
5	Lack of co-operation between entities in the supply chain (B15)	4.00	15	[11,21,23,25,50]
6	Lack of staff empowerment and involvement in LSSI (B11)	4.56	11	[21,23,29,34]
7	Employees and management resistance to process changes (B6)	5.29	6	[1,14,16,22,23,25,26,29,31,35,50]
8	Lack of time for LSSI and its changes (B9)	4.70	9	[14,23,24]
9	Shortage of LSS consultants and trainers (B8)	4.92	8	[14,21,25]
10	Quality problems with the supplied material	3.69	19	[21]
11	Lack of logistic support	3.77	17	[21]
12	Regulations incompatible with LSSI	2.70	25	[50]
13	Lack of communication and transparency across the organization concerning responsibilities during LSSI (B10)	4.68	10	[1,9,20,21,22,25,50]
14	Process performance assessment unsuitable for LSS (lack of suitable KPIs) (B12)	4.43	12	[21,23,25]
15	Uncertain demand and high product variation make it challenging to achieve process stability and capacity for LSSI	3.09	21	[11,25,29]
16	Lack of a specific structured framework for LSSI in CPPs (B7)	5.23	7	[1,9,20,21,22,24,25]
17	Difficulty of process adaptation with technological changes in the industry proposed by LSS	3.04	22	[25,34,50]
18	Evaluation and feedback from industrial employees are not suitable for LSS methodology (B13)	4.25	13	[9,29]
19	Lack of LSS expertise and skills for implementation	3.92	16	[11,14,22,23,24,25,27,29]
20	Lack of training in LSS (B5)	5.46	5	[9,20,21,23,25,29,50]
21	Difficulty in quantifying the benefits of LSS (B14)	4.01	14	[25]
22	Lack of capital in the LSS implementation phase	3.74	18	[1,9,11,14,21,23,24,29,50]
23	High investment and maintenance costs	3.38	20	[14,27]
24	Difficulty in applying Lean in batch production processes, common in the chemical sector	2.94	23	[1]
25	Dependence of (chemical) processes on temperature, chemical reaction time, and sequence makes LSSI challenging	2.60	26	[1]
26	The need to change machines and layout configuration makes LSSI difficult	2.88	24	[9,21,29]

Notes: GM = geometric mean; in bold, the code of the barriers deemed critical for the case of Brazil.

.

In Step 2, through a Delphi survey, the experts determined which barriers they consider critical to LSSI in the CPI in the Brazilian context. They first received an email outlining the study’s objectives and a list of 26 barriers to LSSI. The experts were asked to review the questionnaire for any errors, uncertainties, or overlapping issues. Following corrections, three rounds of the Delphi survey were conducted. In each round, the experts scored the 26 barriers (Table 2) on a 7-point scale ranging from “1—not critical to LSSI” to “7—extremely critical”, with “4—critical” as the midpoint. The geometric mean of the scores was calculated after each round to reduce the influence of extreme values. Barriers with a geometric mean of 4.0 or higher were classified as critical [41]. Nine barriers were identified as critical in the first round. In the second round, after being informed of the results of the first round, the experts identified 15 barriers as critical. In the third round, the experts confirmed that the same 15 critical barriers remained at the top of the rankings (Table 2). Since the ISM model typically recommends focusing on no more than 15 barriers to facilitate contextual relationship analysis [41], no further rounds were conducted. Consensus was considered to have been reached. The 15 critical barriers are presented and described in

Table 3. Critical barriers to LSSI.

ID	Barrier—Description
B1	Lack of systemic understanding of LSS, the need for LSSI, and its benefits—despite its growing popularity in certain industries, LSS remains underutilized in the CPI, leading to challenges in comprehending its concepts, benefits, and the rationale for its implementation [2].
B2	Lack of top management commitment and supervision for LSSI—since managers oversee processes, expenses, and teams, their exemplary dedication and supervision are crucial for adopting necessary changes, as is the case with any process or structural modification. Implementing LSS requires a tailored approach to process and personnel evaluation [51].
B3	Lack of strategic alignment between LSS and corporate strategies for the short and long term—LSS is a dynamic and adaptable approach that can be applied across various industry sectors [50]. However, successful implementation requires alignment with the organization’s infrastructure and corporate strategies. Without this adaptation, LSS is likely to fail [20].
B4	Lack of ownership of LSSI from the staff—leadership must move away from traditional, control-oriented management systems by sharing ownership of LSSI with technicians and operators. This collaborative approach is essential for successful implementation and the widespread adoption of LSS within the organization [21].
B5	Lack of training in LSS—LSSI leads to significant organizational changes. This requires employees to undergo training, receive proper documentation, and study implementation examples to acquire the skills and knowledge necessary for successful LSS adoption [29].
B6	Employees and management resistance to process changes—resistance to change spans all organizational levels and is particularly significant in the case of LSSI [27]. This challenge is aggravated in the CPI, where process changes are often complicated and complex, requiring substantial time and cost investments.
B7	Lack of a specific structured framework for LSSI in the CPI—LSS and its tools were historically developed for assembly-based industries rather than those characterized by continuous processes mixed with batch operations and reliant on physical and chemical variables, as is the case in the CPI [1].
B8	Shortage of LSS consultants and trainers—training is crucial for success when a company is unfamiliar with LSS or is still in the early stages of adoption [52]. External trainers are essential during the initial phases of LSSI to develop and qualify a group of in-house experts. Later, consultants are needed to support the remaining phases of LSSI.
B9	Lack of time for LSSI and its changes—since LSS processes are often not explicitly prioritized, employees tend to view them as secondary tasks, leaving them unable to allocate sufficient time for their study or implementation [27].
B10	Lack of communication and transparency across the organization concerning responsibilities during LSSI—a lack of communication or transparency regarding employee roles and responsibilities during LSSI can lead to failure [14]. Employees need to understand the motivations, goals, and benefits of LSSI.
B11	Lack of staff empowerment and involvement in LSSI—engaging employees at every step of the process is crucial to ensuring that LSS is both implemented and consistently sustained in their daily activities. This involves soliciting their insights, opinions, and concerns while actively empowering them to participate in the process [23].
B12	Process performance assessment not suitable for LSS (lack of suitable KPIs)—once adopted, the metrics for measuring performance in areas such as effectiveness, quality, and productivity must be updated, added, or transformed. Inadequate follow-up on these metrics can lead to incorrect assessments of LSSI success, significantly hindering its effectiveness [37].
B13	Evaluation and feedback from employees are not suitable for LSS—LSSI typically leads to structural changes within the company, introducing new activities, metrics, and daily tasks. The entire employee evaluation process must be reviewed and updated to align with these changes [21].
B14	Difficulty in quantifying the benefits of LSS—LSSI requires a significant initial investment and ongoing expenditures to maintain new structures, consultants, programs, and other elements related to process changes through LSS adoption. This makes it challenging to quantify the benefits [21]. Translating employee skills and product quality improvements into financial indicators to justify the investment is particularly difficult.
B15	Lack of co-operation between entities in the supply chain (suppliers and clients)—LSS requires an initial investment, changes, new implementations, and new ways of working. However, entities often lack the means or willingness to adopt it, leading to misalignments. Without co-operation, these misalignments can create significant challenges for LSSI across the supply chain [11].

.

3.4.2. Stage II: ISM Model

The ISM model development involved the methodological framework’s five steps of Phase II [44]. In Step 1, experts in FGD-1 were asked to conduct pairwise comparisons of the 15 critical barriers (referred to hereafter as “barriers”) to the LSSI. They responded to the following question: “Does barrier i influence or aggravate the impact of barrier j?” Based on their responses, contextual relationships were categorized using four letters:

• “V”—barrier i influences or aggravates the impact of barrier j;
• “A”—barrier j influences or aggravates the impact of barrier i;
• “X”—barriers i and j influence each other;
• “O”—there is no relationship between barriers i and j.

Moderated by one of the authors, the experts progressed from initial individual judgments to a consensus. When consensus was not achieved, the “majority prevailed”.

The contextual relationships among pairs of barriers were recorded in the structural self-interaction matrix (SSIM) (

Table 4. Structural self-interaction matrix (SSIM).

B(i/j)	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15
1	-	V	V	V	O	O	V	O	O	O	O	O	O	V	O
2	-	-	X	V	V	O	O	O	A	V	V	O	O	O	O
3	-	-	-	O	O	V	O	O	O	V	O	V	V	O	O
4	-	-	-	-	V	O	O	V	O	O	V	O	O	V	O
5	-	-	-	-	-	O	V	O	A	X	V	O	O	O	O
6	-	-	-	-	-	-	O	A	O	O	O	V	O	A	A
7	-	-	-	-	-	-	-	O	O	A	O	O	V	V	A
8	-	-	-	-	-	-	-	-	O	A	O	V	O	A	O
9	-	-	-	-	-	-	-	-	-	O	O	O	O	O	O
10	-	-	-	-	-	-	-	-	-	-	O	O	V	O	O
11	-	-	-	-	-	-	-	-	-	-	-	O	O	O	O
12	-	-	-	-	-	-	-	-	-	-	-	-	X	A	O
13	-	-	-	-	-	-	-	-	-	-	-	-	-	O	O
14	-	-	-	-	-	-	-	-	-	-	-	-	-	-	O
15	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-

Note: B(i/j)—barrier in row i or column j.

).

In Step 2, the initial reachability matrix (IRM) results from converting the SSIM into a binary matrix showing the direct relationships among the barriers by substituting the letters with 1 s and 0 s, given the following rules:

• If SSIM(i,j) = “V”, then IRM(i,j) = 1 and IRM(j,i) = 0;
• If SSIM(i,j) = “A”, then IRM(i,j) = 0 and IRM(j,i) = 1;
• If SSIM(i,j) = “X”, then IRM(i,j) = IRM(j,i) = 1;
• If SSIM(i,j) = “O”, then IRM(i,j) = IRM(j,i) = 0.

Finally, the conversion of the IRM into the final reachability matrix (FRM) is obtained through a transitivity check. If barrier i influences barrier j and j influences m, then i indirectly influences m through j, and, if IRM(i,m) is 0, then FRM(i,m) becomes “1*”.

The FRM is partitioned into hierarchical levels in Step 3 through an iterative process. The reachability, antecedent, and intersection sets are generated for each barrier. For barrier i, the reachability set comprises all the barriers that are influenced by it (“1 s” and “1*s” in row i of the FRM), and the antecedent set comprises all the barriers that influence it (“1 s” and “1*s” in column i of the FRM). The intersection set comprises the barriers shared by the antecedent and reachability sets. A barrier i is allocated to the current iteration (hierarchical level) when its reachability and intersection sets match. In practical terms, if barrier B1 influences barriers B2, B3, and B4 (the reachability set of B1) and B2, B3, and B4 (among others in the antecedent set of B1) also influence B1, then B1 is considered to be at its hierarchical level (alongside B2, B3, and B4). Finally, the barriers allocated to a level are separated from the remaining reachability and antecedent sets for the following iterations. The same process is repeated until all the barriers are allocated to a level. In the end, the number of iterations needed corresponds to the hierarchical levels of the ISM model.

In Step 4, the ISM model is established. Initially, a preliminary direct graph is created by arranging barriers vertically based on their hierarchical level and connecting them via the canonical form of the FRM. The indirect links are subsequently removed to obtain the final ISM model. The model illustrates the hierarchical levels and relationships among the barriers to LSSI. In the last step, the experts discuss the ISM model and validate its conceptual consistency.

Finally, in Step 5, the experts in FGD-1 are requested to perform a consistency check of the ISM model. They check the hierarchical structure and interrelations of the barriers to the LSSI, and corrections, if needed, are made in the model. Finally, they are told to verify whether the model correctly represents their “vague” mental model of the system of barriers hindering LSSI [45].

3.4.3. Stage III: MICMAC Analysis

In Stage III, a MICMAC analysis is performed, resulting in a diagram where the barriers are positioned into four clusters based on their driving power (DVP) and dependence (DEP). A barrier’s DVP indicates how many other barriers it influences, whereas its DEP indicates how many barriers influence it. For barrier i, the DVP is determined by the sum of “1” and “1*” in row i of the RFM, while the DEP is determined by the sum of “1” and “1*” in column i of the RFM.

3.4.4. Stage IV: Development of Mitigation Measures

Finally, the MMs for the barriers to the LSSI are developed with input from experts in FGD-2. The MICMAC analysis complements the ISM model and, by merging barriers from the independent cluster (strong DVP and low DEP) with root barriers (highest ISM level), the main barriers to LSSI are identified. These barriers significantly impact the overall system of barriers. Experts must focus on these main barriers when developing MMs, ensuring that the measures impact all the barriers according to their hierarchical relationships, DVPs, and DEPs. To increase the effectiveness of the MMs, the experts should follow the hierarchy outlined in the ISM model, starting with the barriers at the highest level. Nevertheless, dedicated measures to address the barriers in the Autonomous cluster should also be foreseen if necessary because of their reduced DEP. At the end of this stage, experts should verify that at least one MM addresses each barrier. As the MMs consider the hierarchical relationships between barriers, their DVPs, and DEPs and are grounded in expert knowledge and experience, they are expected to be effective and practical.

4. Results

This section presents the results of applying Steps 2 to 5 of Stage II to develop the ISM model and Stage III of the methodological framework to perform the MICMAC analysis.

4.1. ISM Model

In Step 2, Stage II, the SSIM was first converted into the IRM (

Table 5. Initial reachability matrix (IRM).

B(i/j)	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15
1	1	1	1	0	0	0	1	0	0	0	0	0	0	1	0
2	0	1	1	1	1	0	0	0	0	1	1	0	0	0	0
3	0	1	1	0	0	1	0	0	0	1	0	1	1	0	0
4	0	0	0	1	1	0	0	1	0	0	1	0	0	1	0
5	0	0	0	0	1	0	1	0	0	1	1	0	0	0	0
6	0	0	0	0	0	1	0	0	0	0	0	1	0	0	0
7	0	0	0	0	0	0	1	0	0	0	0	0	1	1	0
8	0	0	0	0	0	1	0	1	0	0	0	1	0	0	0
9	0	1	0	0	1	0	0	0	1	0	0	0	0	0	0
10	0	0	0	0	1	0	1	1	0	1	0	0	1	0	0
11	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0
12	0	0	0	0	0	0	0	0	0	0	0	1	1	0	0
13	0	0	0	0	0	0	0	0	0	0	0	1	1	0	0
14	0	0	0	0	0	1	0	1	0	0	0	1	0	1	0
15	0	0	0	0	0	1	1	0	0	0	0	0	0	0	1

Note: B(i/j)—barrier in row i or column j.

). The IRM was subsequently checked for transitivity via an Excel macro, resulting in the FRM (

Table 6. Final reachability matrix (FRM).

B(i/j)	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	DVP
1	1	1	1	1*	1*	1*	1	1*	0	1*	1*	1*	1*	1	0	13
2	0	1	1	1	1	1*	1*	1*	0	1	1	1*	1*	1*	0	12
3	0	1	1	1*	1*	1	1*	1*	0	1	1*	1	1	1*	0	12
4	0	0	0	1	1	1*	1*	1	0	1*	1	1*	1*	1	0	10
5	0	0	0	0	1	1*	1	1*	0	1	1	1*	1*	1*	0	9
6	0	0	0	0	0	1	0	0	0	0	0	1	1*	0	0	3
7	0	0	0	0	0	1*	1	1*	0	0	0	1*	1	1	0	6
8	0	0	0	0	0	1	0	1	0	0	0	1	1*	0	0	4
9	0	1	1*	1*	1	1*	1*	1*	1	1*	1*	1*	1*	1*	0	13
10	0	0	0	0	1	1*	1	1	0	1	1*	1*	1	1*	0	9
11	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	1
12	0	0	0	0	0	0	0	0	0	0	0	1	1	0	0	2
13	0	0	0	0	0	0	0	0	0	0	0	1	1	0	0	2
14	0	0	0	0	0	1	0	1	0	0	0	1	1*	1	0	5
15	0	0	0	0	0	1	1	1*	0	0	0	1*	1*	1*	1	7
DEP	1	4	4	5	7	12	9	11	1	7	8	14	14	10	1

Notes: B(i/j)—barrier in row i or column j; 1*—transitivity; DVP—driving power; DEP—dependence.

). The DVP of each barrier is presented in the last column, and the DEP is in the last row of Table 6.

In Step 3, the FIRM’s level partitioning process requires nine iterations (see the

Table A1. Level partitioning of the final reachability matrix (FRM) iterative process.

Barrier	Reachability Set	Antecedent Set	Intersection Set	Level/ Iteration
1	B: 1; 2; 3; 4; 5; 6; 7; 8; 10; 11; 12; 13; 14	B1	B1
2	B: 2; 3; 4; 5; 6; 7; 8; 10; 11; 12; 13; 14	B: 1; 2; 3; 9	B: 2; 3
3	B: 2; 3; 4; 5; 6; 7; 8; 10; 11; 12; 13; 14	B: 1; 2; 3; 9	B: 2; 3
4	B: 4; 5; 6; 7; 8; 10; 11; 12; 13; 14	B: 1; 2; 3; 4; 9	B4
5	B: 5; 6; 7; 8; 10; 11; 12; 13; 14	B: 1; 2; 3; 4; 5; 9; 10	B: 5; 10
6	B: 6; 12; 13	B: 1; 2; 3; 4; 5; 6; 7; 8; 9; 10; 14; 15	B6
7	B: 6; 7; 8; 12; 13; 14	B: 1; 2; 3; 4; 5; 7; 9; 10; 15	B7
8	B: 6; 8; 12; 13	B: 1; 2; 3; 4; 5; 7; 8; 9; 10; 14; 15	B8
9	B: 2; 3; 4; 5; 6; 7; 8; 9; 10; 11; 12; 13; 14	B9	B9
10	B: 5; 6; 7; 8; 10; 11; 12; 13; 14	B: 1; 2; 3; 4; 5; 9; 10	B: 5; 10
11	B: 11	B: 1; 2; 3; 4; 5; 9; 10; 11	B11	I
12	B: 12; 13	B: 1; 2; 3; 4; 5; 6; 7; 8; 9; 10; 12; 13; 14; 15	B: 12; 13	I
13	B: 12; 13	B: 1; 2; 3; 4; 5; 6; 7; 8; 9; 10; 12; 13; 14; 15	B: 12; 13	I
14	B: 6; 8; 12; 13; 14	B: 1; 2; 3; 4; 5; 7; 9; 10; 14; 15	B14
15	B: 6; 7; 8; 12; 13; 14; 15	B15	B15
1	B: 1; 2; 3; 4; 5; 6; 7; 8; 10; 14	B1	B1
2	B: 2; 3; 4; 5; 6; 7; 8; 10; 14	B: 1; 2; 3; 9	B: 2; 3
3	B: 2; 3; 4; 5; 6; 7; 8; 10; 14	B: 1; 2; 3; 9	B: 2; 3
4	B: 4; 5; 6; 7; 8; 10; 14	B: 1; 2; 3; 4; 9	B4
5	B: 5; 6; 7; 8; 10; 14	B: 1; 2; 3; 4; 5; 9; 10	B: 5; 10
6	B6	B: 1; 2; 3; 4; 5; 6; 7; 8; 9; 10; 14; 15	B6	II
7	B: 6; 7; 8; 14	B: 1; 2; 3; 4; 5; 7; 9; 10; 15	B7
8	B: 6; 8	B: 1; 2; 3; 4; 5; 7; 8; 9; 10; 14; 15	B8
9	B: 2; 3; 4; 5; 6; 7; 8; 9; 10; 14	B9	B9
10	B: 5; 6; 7; 8; 10; 14	B: 1; 2; 3; 4; 5; 9; 10	B: 5; 10
14	B: 6; 8; 14	B: 1; 2; 3; 4; 5; 7; 9; 10; 14; 15	B14
15	B: 6; 7; 8; 14; 15	B15	B15
1	B: 1; 2; 3; 4; 5; 7; 8; 10; 14	B1	B1
2	B: 2; 3; 4; 5; 7; 8; 10; 14	B: 1; 2; 3; 9	B: 2; 3
3	B: 2; 3; 4; 5; 7; 8; 10; 14	B: 1; 2; 3; 9	B: 2; 3
4	B: 4; 5; 7; 8; 10; 14	B: 1; 2; 3; 4; 9	B4
5	B: 5; 7; 8; 10; 14	B: 1; 2; 3; 4; 5; 9; 10	B: 5; 10
7	B: 7; 8; 14	B: 1; 2; 3; 4; 5; 7; 9; 10; 15	B7
8	B8	B: 1; 2; 3; 4; 5; 7; 8; 9; 10; 14; 15	B8	III
9	B: 2; 3; 4; 5; 7; 8; 9; 10; 14	B9	B9
10	B: 5; 7; 8; 10; 14	B: 1; 2; 3; 4; 5; 9; 10	B: 5; 10
14	B: 8; 14	B: 1; 2; 3; 4; 5; 7; 9; 10; 14; 15	B14
15	B: 7; 8; 14; 15	B15	B15
1	B: 1; 2; 3; 4; 5; 7; 10; 14	B1	B1
2	B: 2; 3; 4; 5; 7; 10; 14;	B: 1; 2; 3; 9	B: 2; 3
3	B: 2; 3; 4; 5; 7; 10; 14	B: 1; 2; 3; 9	B: 2; 3
4	B: 4; 5; 7; 10; 14	B: 1; 2; 3; 4; 9	B4
5	B: 5; 7; 10; 14	B: 1; 2; 3; 4; 5; 9; 10	B: 5; 10
7	B: 7; 14	B: 1; 2; 3; 4; 5; 7; 9; 10; 15	B7
9	B: 2; 3; 4; 5; 7; 9; 10; 14	B9	B9
10	B: 5; 7; 10; 14	B: 1; 2; 3; 4; 5; 9; 10	B: 5; 10
14	B14	B: 1; 2; 3; 4; 5; 7; 9; 10; 14; 15	B14	IV
15	B: 7; 14; 15	B15	B15
1	B: 1; 2; 3; 4; 5; 7; 10	B1	B1
2	B: 2; 3; 4; 5; 7; 10	B: 1; 2; 3; 9	B: 2; 3
3	B: 2; 3; 4; 5; 7; 10	B: 1; 2; 3; 9	B: 2; 3
4	B: 4; 5; 7; 10	B: 1; 2; 3; 4; 9	B4
5	B: 5; 7; 10	B: 1; 2; 3; 4; 5; 9; 10	B: 5; 10
7	B7	B: 1; 2; 3; 4; 5; 7; 9; 10; 15	B7	V
9	B: 2; 3; 4; 5; 7; 9; 10	B9	B9
10	B: 5; 7; 10	B: 1; 2; 3; 4; 5; 9; 10	B: 5; 10
15	B: 7; 15	B15	B15
1	B: 1; 2; 3; 4; 5; 10	B1	B1
2	B: 2; 3; 4; 5; 10	B: 1; 2; 3; 9	B: 2; 3
3	B: 2; 3; 4; 5; 10	B: 1; 2; 3; 9	B: 2; 3
4	B: 4; 5; 10	B: 1; 2; 3; 4; 9	B4
5	B: 5; 10	B: 1; 2; 3; 4; 5; 9; 10	B: 5; 10	VI
9	B: 2; 3; 4; 5; 9; 10	B9	B9
10	B: 5; 10	B: 1; 2; 3; 4; 5; 9; 10	B: 5; 10	VI
15	B15	B15	B15	VI
1	B: 1; 2; 3; 4	B1	B1
2	B: 2; 3; 4	B: 1; 2; 3; 9	B: 2; 3
3	B: 2; 3; 4	B: 1; 2; 3; 9	B: 2; 3
4	B4	B: 1; 2; 3; 4; 9	B4	VII
9	B: 2; 3; 4; 9	B9	B9
1	B: 1; 2; 3	B1	B1
2	B: 2; 3	B: 1; 2; 3; 9	B: 2; 3	VIII
3	B: 2; 3	B: 1; 2; 3; 9	B: 2; 3	VIII
9	B: 2; 3; 9	B9	B9
1	B1	B1	B1	IX
9	B9	B9	B9	IX

in Appendix A), resulting in the same number of hierarchical levels, as shown in

Table 7. Level partitioning of the final reachability matrix (FRM).

Barrier	Reachability Set	Antecedent Set	Intersection Set	Level
B11	B11	B: 1; 2; 3; 4; 5; 9; 10; 11	B11	I
B12	B: 12; 13	B: 1; 2; 3; 4; 5; 6; 7; 8; 9; 10; 12; 13; 14; 15	B: 12; 13	I
B13	B: 12; 13	B: 1; 2; 3; 4; 5; 6; 7; 8; 9; 10; 12; 13; 14; 15	B: 12; 13	I
B6	B6	B: 1; 2; 3; 4; 5; 6; 7; 8; 9; 10; 14; 15	B6	II
B8	B8	B: 1; 2; 3; 4; 5; 7; 8; 9; 10; 14; 15	B8	III
B14	B14	B: 1; 2; 3; 4; 5; 7; 9; 10; 14; 15	B14	IV
B7	B7	B: 1; 2; 3; 4; 5; 7; 9; 10; 15	B7	V
B5	B: 5; 10	B: 1; 2; 3; 4; 5; 9; 10	B: 5; 10	VI
B10	B: 5; 10	B: 1; 2; 3; 4; 5; 9; 10	B: 5; 10	VI
B15	B: 15	B15	B15	VI
B4	B4	B: 1; 2; 3; 4; 9	B4	VII
B2	B: 2; 3	B: 1; 2; 3; 9	B: 2; 3	VIII
B3	B: 2; 3	B: 1; 2; 3; 9	B: 2; 3	VIII
B1	B1	B1	B1	IX
B9	B9	B9	B9	IX

.

The ISM model was established in Step 4 with the help of the canonical form of the FRM (

Table 8. Canonical form of the RFM.

B(i/j)	B11	B12	B13	B6	B8	B14	B7	B5	B10	B15	B4	B2	B3	B1	B9
B11	I	0	0	0	0	0	0	0	0	0	0	0	0	0	0
B12	0	I	1	0	0	0	0	0	0	0	0	0	0	0	0
B13	0	1	I	0	0	0	0	0	0	0	0	0	0	0	0
B6	0	1	1	II	0	0	0	0	0	0	0	0	0	0	0
B8	0	1	1	1	III	0	0	0	0	0	0	0	0	0	0
B14	0	1	1	1	1	IV	0	0	0	0	0	0	0	0	0
B7	0	1	1	1	1	1	V	0	0	0	0	0	0	0	0
B5	1	1	1	1	1	1	1	VI	1	0	0	0	0	0	0
B10	1	1	1	1	1	1	1	1	VI	0	0	0	0	0	0
B15	0	1	1	1	1	1	1	0	0	VI	0	0	0	0	0
B4	1	1	1	1	1	1	1	1	1	0	VII	0	0	0	0
B2	1	1	1	1	1	1	1	1	1	0	1	VIII	1	0	0
B3	1	1	1	1	1	1	1	1	1	0	1	1	VIII	0	0
B1	1	1	1	1	1	1	1	1	1	0	1	1	1	IX	0
B9	1	1	1	1	1	1	1	1	1	0	1	1	1	0	IX

Notes: B(i/j)—barrier in row i or column j; the hierarchical levels are displayed on the diagonal of the matrix; the barriers at the same level are shaded in gray.

). Finally, in Step 5, the experts discussed the ISM model. They validated its conceptual consistency, recognizing that it effectively captured their “vague” mental representation of the system of barriers hindering the LSSI in the Brazilian CPI. The ISM model is shown in Figure 2.

4.2. MICMAC Analysis

Following Step III of the methodological framework, a driving power-dependence diagram for the MICMAC analysis is created using the DVP and DEP values found in Table 6 (Figure 3).

5. Discussion and Mitigation Measures

This section presents the findings from FGD-2. First, the experts discussed the ISM model and MICMAC analysis. They then proposed MMs to address the barriers to LSSI.

5.1. Discussion

Although it is not within the scope of this study to analyze the ranking of barriers to LSSI (Table 2), two main findings emerge. First, the top 15 barriers, ranked by criticality, are consistent with those in similar studies, albeit not in the same order due to varying contexts [5]. Second, notable process-industry-specific barriers, such as “difficulty in applying Lean in batch production processes, common in the chemical sector” and “(chemical) process dependence on temperature, chemical reaction time, and sequence making LSSI challenging”, are ranked 23rd and 26th, respectively, indicating that they are not critical. This contradicts the findings of Panwar et al. [1]. Their research was based on the opinions of Indian process industry managers. These managers strongly believed that the need “to produce in large batches for proper utilization of capacity and equipment” was a significant reason for not adopting lean manufacturing to increase performance. In contrast, the experts in the present study were well versed in LSS and CPI and did not view these barriers as the most critical to LSSI.

Concerning the ISM model (Figure 2), Level I is the lowest hierarchical level and comprises B11, B12, and B13, with B12 and B13 influencing each other. Level II includes only B6, which directly influences B12 and B13. If they do not embrace the changes associated with the LSS, then the performance assessment and their feedback are unlikely to be adequate.

Level III contains only B8, which influences B6. Insufficient trainers during the initial phases of LSSI, as well as a lack of consultants in later stages, hinder the necessary changes that employees and managers must undergo to implement and maintain LSS successfully [21]. B14 at Level IV influences B8. Difficulties in measuring the benefits of LSS amplify the impact of the shortage of consultants and trainers. B7 at Level V influences B14. The lack of a specific framework for implementing LSS makes it a challenge to evaluate and quantify the benefits of LSS [1].

Level VI comprises B5, B10, and B15, which influence B7. The experts agree that those barriers, related to a lack of training, communication, transparency, and co-operation, do not help create a properly structured framework for LSSI in the CPI. In turn, a lack of training, communication, and transparency (B5 and B10), which are factors that influence each other, contributes to reduced staff empowerment and involvement in LSSI (B11 at Level I) [23]. Notably, although B15 is at an intermediate level, it is not influenced by any hierarchically superior barrier. Level VII includes B4, which influences B5 and B10. The experts highlighted that the lack of staff ownership of LSSI exacerbates the impact of a lack of training, organizational communication, and transparency in LSSI. B3 and B2 comprise Level VIII, which influence each other and B4. The lack of strategic alignment among LSSs and co-operative strategies and top managers’ commitment to and supervision of LSSI influence each other, and both influence the lack of staff ownership of LSSI [25].

Level IX, the highest hierarchal level at the basis of the ISM model, comprises B1 and B9, which are deemed the root barriers to the LSSI in the CPI. These barriers do not influence each other but directly influence B2 and B3. The lack of understanding of LSS, its benefits and time for its implementation, and its changes increase the impacts of strategic misalignment and the lack of top management commitment and supervision [24]. Moreover, B1 and B9 indirectly influence all the other barriers except B15.

In the MIMAC analysis (Figure 3), the Dependent cluster comprises B6, B7, B8, B12, B13, and B14. Owing to their weak DVP and high DEP, these barriers are strongly influenced but have little or no influence on the other barriers. The Linkage cluster has no barriers, which has been previously found in the literature [45]. Barriers in this cluster exhibit strong DVP and high DEP, indicating relationships between barriers in the Independent and Dependent clusters. The Autonomous cluster includes only B15; its low DEP indicates that other barriers do not significantly influence it. However, in the present study, its intermediate DVP suggests that it may influence other barriers. Finally, the Independent cluster comprises B1 to B5, B9, and B10. Owing to their strong DVP and low DEP, these barriers influence the other barriers, except B15, and are not influenced by them. According to the MICMAC analysis, these are the key barriers that hamper the implementation of LSS. Owing to the absence of barriers in the Linkage cluster, these barriers directly influence those in the Dependent cluster, as evidenced in the ISM model.

As expected, the results of the MICMAC analysis complement those of the ISM model and complement each other. The MICMAC analysis identifies barriers with stronger DVPs found in the Independent cluster (B1, B9, B3, B2, B4, B5, and B10); it also identifies barriers with higher DEPs in the Dependent cluster (B6, B7, B8, B12, B13, and B14). The ISM model adds depth to this information by demonstrating the hierarchal relationships among these barriers. Moreover, the MICMAC analysis’s indication that B15 is in the autonomous cluster is complemented by the ISM model, which indicates that it influences all lower hierarchical barriers (B7, B14, B8, B6, B12, B13, and B11), as previously predicted. Understanding this is crucial for creating effective MMs that address barriers in the Independent cluster and, through hierarchical interrelations, address barriers in the Dependent cluster. MMs must be designed with this understanding to ensure their effectiveness.

5.2. Mitigation Measures

The FGD-2 experts were tasked with developing MMs for barriers to LSSI. The moderator started by educating the experts that the MMs should act on the main barriers (the ISM model root barriers and the MICMAC analysis Independent barriers), namely, B1 to B5, B9, and B10. The moderator emphasized the importance of following the hierarchy outlined in the ISM model and beginning the development of MMs by addressing the barriers at the highest levels (B1 and B9 at Level IX). Furthermore, these MMs must effectively reach and influence the remaining barriers according to the hierarchical relationships outlined in the ISM model. The experts were also informed that dedicated MMs to B15, a barrier in the Autonomous cluster, should be planned if necessary. Finally, they were asked to check whether at least one MM impacted each barrier.

During the FGD-2 session, to understand, clarify, and validate the effectiveness of the MMs proposed by the experts, the moderator used open questions such as “How do you explain this MM?”, “Can you provide more details?”, “How can this MM be improved?”, “How can these (similar) MMs be merged?”, or “Based on your experience, can you give an example of how this measure mitigates this barrier?”

At the end of the FGD-2 session, the experts reached a consensus and proposed 10 MMs. To mitigate B1, they recommended “Creating an organization-wide training program to increase awareness of LSS” (MM1). This training program should address the transversal nature of LSS methodologies and clarify the roles of managers, technicians, and operators. Additionally, it should highlight sustainable awareness and management practices such as energy efficiency, waste management, water conservation, and carbon management and integrate insights from external LSS experts and chemical industry experts. The experts believe that the training would impact B2, B3, B4, and B10 [10].

The proposal to address B9 is the “Creation of a multiarea mission group for LSSI” (MM2). As suggested by two experts, this group must be formed by a leader and at least one member from each area. This group’s mission should be to ensure proper LSSI by qualifying members, tracking initiatives, and promptly sharing valuable information throughout the organization.

The third measure (MM3) proposes “Mandatory certification for mission group members by external experts”. Members must complete a specified number of training hours annually, with periodic certification validation.

Another measure suggested for B2 is the “Institutionalization of promotion and bonus packages for managers most active and influential in LSSI” (MM4). The experts suggested extending this measure to include technicians and operators. This would allow employees to enhance their role within the company, become champions of LSS training and implementation, encourage others, and promote cultural change [25].

The experts recommended the “Implementation of an annual competition to award best practices in LSSI and innovation in the company” (MM5), which aligns with the findings of Jadhav et al. [21]. They also emphasized the importance of promoting practices that encourage the adoption of sustainable initiatives. This measure directly impacts B4, among other barriers.

“Creating an online repository accessible to all employees with information on LSS concepts, tools, and key contacts” (MM6) would directly impact B10 and help mitigate other barriers. This measure supports the shift to paperless operations, which use documentation and digital tools to monitor and optimize resource usage. Additionally, providing controlled access to employees of other entities in the supply chain could enhance co-operation (B15).

Although MM6 impacts B15, owing to the barrier’s autonomous nature, experts have proposed “Developing specific training programs to help suppliers and customers implement LSS and improve their interface with the company’s operations, along with new relationship metrics aligned with LSS” (MM7), in line with Leite et al. [11]. These programs should also encourage suppliers and customers to adopt sustainable practices.

Finally, experts proposed “Establishing centers within Brazilian chemical industry associations to implement and maintain LSS” (MM8). These centers would promote LSS through roadshows, workshops, seminars, and exhibitions and serve as repositories of best practices, offering training courses and expert resources [24]. Additionally, they suggested another similar measure, “Creating centers with Brazilian Lean and LSS associations specialized in LSSI in the CPI” (MM9) and “Incorporating LSS education and training into university programs related to the CPI” (MM10). University education and training programs were recommended to incorporate basic sustainability concepts and practices. For the remaining dependent barriers, the experts concluded that the proposed measures indirectly address them; thus, no additional measures were needed.

Finally, given the significance of the CPI in the economy, the BOP, and the industry’s overall sustainability, the Brazilian government should support measures MM8 and MM9 and encourage universities to implement MM10. Experts recommended that this can be achieved through financial incentives and dynamization programs.

6. Conclusions

Using the case of Brazil, this study focused on researching the main barriers to LSSI in the CPI in an emergent economy and proposed respective MMs. The successful implementation of LSS contributes to the sustainability of management practices being more sustainable and, thus, to that of the industry overall. The adopted methodology is grounded in an ISM-MICMAC analysis supported by a literature review, a Delphi survey, and two focus group discussions.

Through the Delphi survey, 26 barriers from the literature were ranked in terms of criticality to LSSI in the CPI. Surprisingly, industry-specific barriers, such as the “challenges in applying Lean methodologies in batch production” and the “process reliance on temperature, chemical reaction time, and sequence”, which are prevalent in the CPI, are not listed among the top 15. These findings are unexpected and contradict the literature.

The main barriers to LSSI stem from the ISM model and MICMAC analysis. They include (i) a lack of systemic understanding of LSSI and its benefits, (ii) insufficient top management commitment and supervision, (iii) strategic misalignment between LSS and corporate strategies, (iv) limited staff ownership of LSSI, (v) inadequate training, (vi) insufficient time allocation for the LSSI and changes, and (vii) poor communication and transparency across the organization concerning responsibilities during LSSI. These barriers govern the dynamics of the overall barrier system. Therefore, effectively mitigating this system of barriers must be grounded in MMs that act on and address these barriers and can also address the remaining barriers.

The experts proposed 10 MMs, which were divided into two groups. The first includes organization-level measures such as (i) developing company-wide LSS training programs, (ii) establishing an LSSI mission group, (iii) mandatory certification, (iv) bonus packages for managers active in LSSI, (v) awarding best practices in LSSI and innovation, and (vi) creating a comprehensive online repository for LSS information. The second covers CPI-specific measures, such as (vii) developing specific training programs to help extend LSSI to suppliers and customers, (viii) establishing centers within chemical industry associations to implement and maintain LSS, (ix) creating centers with Lean and LSS associations specialized in LSSI in the CPI, and (x) incorporating LSS education and training into university programs. The experts highlighted the government’s role in the last three measures, either through financial support or promotion.

This study makes two primary contributions. First, it contributes to the literature by proposing an innovative methodology to identify the main barriers to LSSI in the CPI and to develop measures grounded in the relationships between barriers that are capable of systematically mitigating them. Second, the methodology was successfully applied to the context of Brazil, a developing economy. Supporting LSSI in the Brazilian CPI would promote sustainability in the industry by enhancing operational efficiency through waste reduction, lowering transportation-related emissions due to decreasing chemical product imports, and fostering economic growth and job creation.

There are also practical implications for the CPI regarding LSSI. Project managers must adopt a holistic approach beyond individual barriers and their relative importance when designing MMs. To effectively address the impact of barriers to LSSI, managers should (1) evaluate the significance of barriers based on their interrelations, hierarchy, driving power, and dependence; (2) recognize that, while MMs must target primary barriers, they should also influence others; and (3) ultimately, support the process with expert opinions, combining individual professional experiences and group interactions. Moreover, experts appreciated the practicality and value of this approach for developing MMs, as it offers a structured framework for their development.

This study has several limitations. First, the FGDs used to develop the ISM model and the MICMAC analysis considered only 15 of the 26 barriers listed in the initial questionnaire for ranking. However, the barriers’ rankings do not always reflect the hierarchical position of the barriers in the ISM model or their power scores in the MICMAC analysis. By including more barriers in the FGDs, we might have discovered other lower-ranking barriers that could have been considered among the primary ones. Second, the findings obtained for the CPI in the Brazilian context may not be generalizable to other contexts; nevertheless, the proposed methodology can. Third, the developed MMs depend on expert subjectivity, necessitating objective effectiveness testing. However, these limitations constitute future research opportunities. The relationships between barriers can be quantitatively verified via appropriate statistical analysis tools such as structural equation modeling (SEM). Furthermore, the proposed methodology can be replicated in different geographic contexts.