A new hybrid method for manufacturing sustainability performance assessment: a case study in furniture industry

ABSTRACT

Manufacturing sustainability performance evaluation is critical for improving the production line’s performance. Three aspects involved in manufacturing sustainability are economic, social, and environmental. This research aims to create a new framework for assessing the Manufacturing Sustainability Score (MSS) based on the lean concept. A hybrid method involving the Delphi, Dematel-Analytical Network Process (ANP), Sustainability Value Stream Mapping (VSM), and Traffic Light System is proposed. The proposed framework is applied to a case study in the furniture industry, and the results demonstrate that the new framework can assess manufacturing sustainability performance. A case study’s manufacturing sustainability score is 66.97 percent. It demonstrates that the company’s performance is moderate and could be improved. The managerial and academic implications are also presented in this study.

KEYWORDS:

• Manufacturing sustainability
• sustainable manufacturing
• lean manufacturing
• value stream mapping (VSM)

1. Introduction

Population growth, resource depletion, and environmental degradation have all become significant challenges for humans in recent decades (Zhu et al., 2020). The most common environmental issues are resource depletion, increased pollution, and ecological degradation (Sultan et al., 2021; Utama, Widodo, Ibrahim, Dewi et al., 2020). It has prompted businesses to incorporate environmental management concerns into their strategic plans (J. K. Y. J. K. Y. Lee et al., 2021). Waste, toxic chemicals, and greenhouse gas emissions from the manufacturing process all contribute to the manufacturing industry’s environmental impact (Ibrahim et al., 2020; Mohammad Ebrahimi & Koh, 2021). This issue causes stakeholders to be concerned about environmental issues (Helleno et al., 2017; Utama et al., 2022). As a result, eliminating environmental pollution and achieving sustainable development is a corporate social responsibility (Simpson, 2012). The concept of sustainable manufacturing refers to a manufacturing system process that takes into account economic, environmental, and social factors (Machado et al., 2020). This concept aims to produce products in a sustainable manner while maintaining global competitiveness and facing the most recent challenges and problems (J. Y. J. Y. Lee et al., 2014). This concept necessitates the use of more environmentally friendly processes and materials in the production of sustainable products, which is an essential component of sustainable manufacturing (Pusavec et al., 2010). Manufacturing sustainability should be assessed to ensure that sustainable manufacturing is achieved (Cherrafi et al., 2016). Therefore, businesses must abandon traditional techniques that only emphasize cost reduction and efficiency improvement without considering the environmental and social impacts of their operations (Faulkner & Badurdeen, 2014).

Some researchers have contributed to corporate sustainability performance measurement (Searcy, 2012). The researchers proposed framework is used for corporate reporting (Moldavska & Welo, 2019). However, this framework does not generate useful information for shop-floor decision-makers at the internal and production line levels. Currently, research into manufacturing sustainability studies is rarely conducted on the shop floor (Patalas-Maliszewska et al., 2022; Swarnakar et al., 2021). Previous research suggests a conceptual manufacturing sustainability assessment framework (Swarnakar et al., 2022). However, it is difficult to apply in industries with varying characteristics. As a result, in order to measure sustainability performance at the manufacturer level, different frameworks and indicators for assessing manufacturing sustainability must be considered from a technical standpoint (Theng et al., 2021). J. Y. J. Y. Lee et al. (2014) provided a shop-floor index based on engineering principles to measure a facility’s sustainable performance. Huang and Badurdeen (2018) proposed a similar method for evaluating the success of sustainable manufacturing in a production line. However, due to a lack of research, the mapping of sustainable manufacturing in the production line is not presented. In fact, visualizing the production process at each work station assists decision-makers in making better sustainability decisions (Sri Hartini et al., 2017b). Hence, a new methodology is required to map and measure sustainable manufacturing performance at the production line level based on economic, social, and environmental variables.

Value stream mapping (VSM) is a mapping tool used by manufacturers to track and improve production line performance (Theng et al., 2021; Utama et al., 2016). VSM determines whether or not an activity is value-added at each stage of production. Traditional VSM does not consider environmental and social issues when mapping the production process (Dewi et al., 2021; Sri; Hartini et al., 2017b). Instead, it focuses on mapping economic aspects such as inventory, time, transportation, and defects (Schoeman et al., 2021). As a result, conventional VSM does not take into account sustainability characteristics such as environmental and social aspects of manufacturing companies. Brown et al. (2014) created the Sustainable-VSM framework for mapping on the production line. Several other researchers use sustainable-VSM in the furniture manufacturing industry (Sri Hartini et al.), electronics (Faulkner et al.) (Sri Hartini et al.), manufacturing (Mishra et al., 2020), and textiles (S. Hartini et al., 2021) to map performance in each workstation and process. Unfortunately, the sustainable VSM tool cannot calculate the overall manufacturing sustainability performance index. The total performance score in sustainable manufacturing must be measured precisely using appropriate indicators. For this reason, Sri Hartini, Ciptomulyono, Anityasari et al. (2020) proposed a methodology for measuring the new manufacturing sustainability index based on appropriate indicators and indicator weights based on sustainable VSM mapping. To our knowledge, only Sri Hartini, Ciptomulyono, Anityasari et al. (2020) research has used sustainable VSM to measure manufacturing sustainability scores based on indicator weights. This study, however, ignores the relationship between indicators. Each indicator has a relationship that influences the indicator’s weight (Liu et al., 2021; Quezada et al., 2018). Therefore, when evaluating performance, the relationship between the indicators of manufacturing sustainability must be investigated.

Based on the gaps and research opportunities described in the previous paragraph, Manufacturing Sustainability Performance Assessment (MSPA) needs to consider the relationship between manufacturing sustainability indicators. Therefore, in this study, the research goals (RG) are as follows:

RG 1. Develop a new MSPA framework by considering the interrelationships between manufacturing sustainability indicators.

RG 2. Evaluating the performance of manufacturing sustainability based on case studies in furniture industry by considering the relationship between manufacturing sustainability indicators.

Based on these objectives, a new framework is proposed, starting with the identification of indicators and the determination of relevant indicators. Furthermore, an assessment of indicator linkage is determined, followed by indicator weighting based on the relationship indicator. The overall manufacturing sustainability score assessment is based on sustainable VSM mapping and indicator weighting. A case study in the furniture industry was also used to evaluate the performance of manufacturing sustainability. Therefore, it is clear that this research makes a real contribution to science by providing a new framework for assessing manufacturing sustainability performance that considers the interrelationships between manufacturing sustainability indicators. Furthermore, this study also evaluates the furniture industry’s manufacturing sustainability performance.

The remainder of this article is as follows: The second section is a review of the literature. Section 3 introduces the proposed MSPA framework. Section 4 discusses how the proposed framework was applied to a case study. Section 5 describes the results, discussion, and practical managerial implications. This article concludes with a conclusion and future research.

2. Literature review

This section provides a literature review based on previous research on manufacturing sustainability assessment. Elkington (1998) proposed a framework for assessing sustainability that is based on the Triple Bottom Line (TBL). TBL is a framework for measuring an organization’s performance by taking three factors into account: economic, social, and environmental. Rogers and Hudson (2011) refer to this factor as an indicator of sustainable development.

Appropriate TBL sustainability indicators are critical in assessing manufacturing sustainability (Saavalainen et al., 2017). However, inappropriate indicators can provide operational managers with less useful information regarding the performance of the manufacturing process. Table 1 summarizes some research on MSPA. It demonstrates that MSPA research is dominated by production flow mapping and MSPA score assessment, which were investigated separately. MSPA, on the other hand, can be evaluated using production flow mapping (Sri Hartini et al.). Furthermore, previous research has a tendency to select indicators based on theoretical indicators. As a result, some indicators become less relevant to the case study issue. Furthermore, previous studies ignored the interdependence of TBL indicators. As a result, it affects the weight of the TBL indicator used for MSPA, which is less important. To provide appropriate information to operations managers, the interrelationships between TBL indicators must be considered in order to assess sustainable manufacturing performance.

Table 1. Summary of Manufacturing Sustainability Assessment Research.

Study	Rating Score	Mapping	Indicator selection process	Interrelation indicators	Indicator	Aplications	Tools
Brown et al. (2014)	-	V	Theoretical	-	TBL	Electronic	VSM
Castiglione et al. (2022)	V	-	Theoretical	-	Economic, Enviromental	Manufacturing	MEIO
Paju et al. (2010)	-	V	Theoretical	-	Enviromental	Not Clear	LCA, VSM
Bhadu et al. (2022)	V	-	Empirical	-	Economic	Manufacturing	AHP
Faulkner and Badurdeen (2014)	-	V	Theoretical	-	TBL	Electronic	VSM
Helleno et al. (2017)	-	V	Theoretical	-	TBL	Cosmetic	VSM
Sri Hartini et al.	-	V	Empirical	-	TBL	Furniture	VSM
Huang and Badurdeen (2018)	V	-	Theoretical	-	TBL	Electronic	AHP
Garza-Reyes, Kumar et al. (2018)	-	V	Theoretical	-	TBL	Mining	VSM
Bait et al. (2020)	-	V	Theoretical	-	TBL	Manufacturing	Simulation, VSM
Swarnakar et al. (2021)	-	V	Empirical	-	TBL	Manufacturing	VSM
Patalas-Maliszewska et al. (2022)	V	-	Theoretical	-	TBL	Manufacturing	Fuzzy AHP, Decision Tree
Soltani et al. (2020)	V	-	Theoretical	-	TBL	Gas bottle	AHP, VSM, TOPSIS,
Sri Hartini, Ciptomulyono, Anityasari et al. (2020)	V	V	Empirical	-	TBL	Furniture	Delphi, AHP, VSM
This Research	V	V	Empirical	V	TBL	Furniture	Delphi, Dematel, ANP,VSM

Another gap in MSPA is that this problem is widely used in the manufacturing and electronics fields. However, this issue is still understudied in the furniture industry. To fill the void, this study proposes MSPA based on process flow mapping in the furniture industry, which includes interrelationships between TBL indicators. Based on empirical studies, this study proposes the TBL indicator used in this study. Furthermore, the manufacturing sustainability rating score is determined by the mapping results of each TBL indicator in each process.

The Delphi method is proposed as a method for identifying relevant indicators. The Delphi method is simple to use to evaluate the relevance of indicators and criteria applied to various sustainable development issues, such as MSPA (Sri Hartini, Ciptomulyono, Anityasari et al., 2020) and sustainable supplier selection (Cherier & Meliani, 2019). The Dematel-ANP procedure was proposed to evaluate the interrelationship and indicator weighting of indicators. Dematel-ANP is a popular method for assessing the interrelation of indicators and indicator weighting in a variety of issues, such as green supplier selection (Utama, Maharani et al., 2021; Utama, Putri et al., 2021), sustainable supply chain performance (Leksono et al., 2019; Rao, 2021), and mapping strategy (Ocampo et al., 2020). In addition, sustainable VSM was selected because it is a comprehensive procedure for mapping production line indicators (J. K. Y. J. K. Y. Lee et al., 2021; Soltani et al., 2020). In the following section, the MSPA framework is described in detail.

3. Framework proposed MSPA

In this section, the proposed MSPA framework is proposed to assess the manufacturing sustainability score. Figure 1 shows the proposed MSPA framework. The following subsection describes the five major stages of MSPA.

Figure 1. Proposed MSPA Framework.

3.1. Indicator collection and indicator formula

The proposed framework’s initial stage begins with the compilation of efficiency indicators and formulas for each TBL indicator. The indicators are gathered by conducting a literature review on manufacturing sustainability in the production line. Table 2 shows the manufacturing sustainability indicators that emerged from the literature review. Table 3 also contains the performance formula for each indicator.

Table 2. Manufacturing sustainability indicators based on literature review.

Factors	Indicators	References
Economic	Cost	(Hartini, Ciptomulyono, Anityasari et al., 2020; Helleno et al., 2017; Torres & Gati, 2009; Vinodh et al., 2014)
	Quality	(Hartini, Ciptomulyono, Anityasari et al., 2020; Sparks, 2014)
	Invetory	(Atieh et al., 2016; Faulkner & Badurdeen, 2014; Hartini, Ciptomulyono, Anityasari et al., 2020; Helleno et al., 2017; Sparks, 2014; Vinodh et al., 2014)
	Time	(Faulkner & Badurdeen, 2014; Fearne & Norton, 2006; Garza-Reyes, Romero et al., 2018; Sri Hartini et al., 2017a; Hartini, Ciptomulyono, Anityasari et al., 2020; Paju et al., 2010)
Enviroment	Energy	(Hartini, Ciptomulyono, Anityasari et al., 2020; Helleno et al., 2017; Torres & Gati, 2009; Vinodh et al., 2014)
	Material	(Hartini, Ciptomulyono, Anityasari et al., 2020; Helleno et al., 2017; Torres & Gati, 2009; Vinodh et al., 2014)
	Waste Recycle	(Hartini, Ciptomulyono, Anityasari et al., 2020)
Social	Employee training	(Hartini, Ciptomulyono, Anityasari et al., 2020)
	Safety level	(Hartini, Ciptomulyono, Anityasari et al., 2020; Helleno et al., 2017)
	Health level	(Hartini, Ciptomulyono, Anityasari et al., 2020)
	Satisfaction level	(Hartini, Ciptomulyono, Anityasari et al., 2020)

Table 3. Performance formulas for each indicator.

No. (i)	Indicators	Input	Formula	References
1	Cost (Rp)	$\mathrm{C}\mathrm{E}$ = Cost efficiency $\mathrm{V}\mathrm{A}\mathrm{C}$ = Cost of activities that have added value $\mathrm{N}\mathrm{V}\mathrm{A}\mathrm{C}$ = Cost of non-value added activities $\mathrm{T}\mathrm{C}$ = Total cost $\mathrm{n}$ = process to $\mathrm{n}$	$\mathrm{C}\mathrm{E}=\frac{\mathrm{V}\mathrm{A}\mathrm{C}}{\mathrm{T}\mathrm{C}}$$\mathrm{V}\mathrm{A}\mathrm{C}=\sum _{\mathrm{i}=1}^{\mathrm{n}}(\mathrm{V}\mathrm{A}\mathrm{C}\mathrm{i})$$\mathrm{N}\mathrm{V}\mathrm{A}\mathrm{C}=\sum _{\mathrm{i}=1}^{\mathrm{n}}(\mathrm{N}\mathrm{V}\mathrm{A}\mathrm{C}\mathrm{i})$$\mathrm{T}\mathrm{C}=\mathrm{V}\mathrm{A}\mathrm{C}+\mathrm{N}\mathrm{V}\mathrm{A}\mathrm{C}$	Vinodh et al. (2014)
2	Quality (part)	QE = Quality efficiency ND = Number of defects TM = Total material	$\mathrm{Q}\mathrm{E}=1-\left(\frac{\mathrm{N}\mathrm{D}}{\mathrm{T}\mathrm{M}}\right)$	Sparks (2014)
3	Inventory (unit)	$\mathrm{I}\mathrm{E}$ = Inventory efficiency $\mathrm{N}\mathrm{I}$ = Total inventory $\mathrm{T}\mathrm{M}$ = Total material	$\mathrm{I}\mathrm{E}=\frac{\mathrm{N}\mathrm{I}}{\mathrm{T}\mathrm{M}}$	Faulkner and Badurdeen (2014)
4	Time (menit)	$\mathrm{T}\mathrm{E}$ = Time efficiency $\mathrm{V}\mathrm{A}\mathrm{T}$= Time in value-added activities $\mathrm{T}\mathrm{T}$ = Total time $\mathrm{N}\mathrm{V}\mathrm{A}\mathrm{T}$ = Time in non-value added activities $\mathrm{n}$ = process to $\mathrm{n}$	$\mathrm{T}\mathrm{E}=\frac{\mathrm{V}\mathrm{A}\mathrm{T}}{\mathrm{T}\mathrm{T}}$$\mathrm{V}\mathrm{A}\mathrm{T}=\sum _{\mathrm{i}=1}^{\mathrm{n}}(\mathrm{V}\mathrm{A}\mathrm{T}\mathrm{i})$$\mathrm{N}\mathrm{V}\mathrm{A}\mathrm{T}=\sum _{\mathrm{i}=1}^{\mathrm{n}}(\mathrm{N}\mathrm{V}\mathrm{A}\mathrm{T}\mathrm{i})$$\mathrm{T}\mathrm{T}=\mathrm{V}\mathrm{A}\mathrm{T}+\mathrm{N}\mathrm{V}\mathrm{A}\mathrm{T}$	Vinodh et al. (2014)
5	Energy (kWh)	EE = Efisiensi energi VAE = energi pada kegiatan yang memiliki nilai tambah NVAE = energi pada kegiatan yang tidak memiliki nilai tambah ET = energi total $\mathrm{n}$ = process to $\mathrm{n}$	$\mathrm{E}\mathrm{E}=\frac{\mathrm{V}\mathrm{A}\mathrm{E}}{\mathrm{T}\mathrm{E}}$$\mathrm{V}\mathrm{A}\mathrm{E}=\sum _{\mathrm{i}=1}^{\mathrm{n}}(\mathrm{V}\mathrm{A}\mathrm{E}\mathrm{i})$$\mathrm{N}\mathrm{V}\mathrm{A}\mathrm{E}=\sum _{\mathrm{i}=1}^{\mathrm{n}}(\mathrm{N}\mathrm{V}\mathrm{A}\mathrm{E}\mathrm{i})$$\mathrm{T}\mathrm{E}=\mathrm{V}\mathrm{A}\mathrm{E}+\mathrm{N}\mathrm{V}\mathrm{A}\mathrm{E}$	Vinodh et al. (2014)
6	Material (kg)	$\mathrm{M}\mathrm{E}$ = Material efficiency $\mathrm{V}\mathrm{A}\mathrm{M}$ = material on activities that have added value $\mathrm{T}\mathrm{M}$ = total material used $\mathrm{N}\mathrm{V}\mathrm{A}\mathrm{M}$ = material on activities that do not have added value $\mathrm{n}$ = process to $\mathrm{n}$	$\mathrm{M}\mathrm{E}=\frac{\mathrm{V}\mathrm{A}\mathrm{M}}{\mathrm{T}\mathrm{M}}$$\mathrm{V}\mathrm{A}\mathrm{M}=\sum _{\mathrm{i}=1}^{\mathrm{n}}(\mathrm{V}\mathrm{A}\mathrm{M}\mathrm{i})$$\mathrm{N}\mathrm{V}\mathrm{A}\mathrm{M}=\sum _{\mathrm{i}=1}^{\mathrm{n}}(\mathrm{N}\mathrm{V}\mathrm{A}\mathrm{M}\mathrm{i})$$\mathrm{T}\mathrm{M}=\mathrm{V}\mathrm{A}\mathrm{M}+\mathrm{N}\mathrm{V}\mathrm{A}\mathrm{M}$	Vinodh et al. (2014)
7	Waste Recycle (kg)	$\mathrm{T}\mathrm{W}$ = Total waste $\mathrm{W}\mathrm{L}$ = Amount of land waste	$\mathrm{W}\mathrm{E}=1-\left(\frac{\mathrm{W}\mathrm{L}}{\mathrm{T}\mathrm{W}}\right)$	Helleno et al. (2017)
8	Employee training	${\mathrm{E}}_{\mathrm{H}\mathrm{R}\mathrm{D}}$ = Employee training level $\mathrm{N}\mathrm{T}$ = Number of employees who attended the training $\mathrm{N}\mathrm{E}$ = Total number of employees	${\mathrm{E}}_{\mathrm{H}\mathrm{R}\mathrm{D}}=\frac{\mathrm{N}\mathrm{T}}{\mathrm{N}\mathrm{E}}$	Hartini, Ciptomulyono, Anityasari et al. (2020)
9	Safety level	$\mathrm{N}\mathrm{R}$ = Number of activities at risk $\mathrm{N}\mathrm{a}\mathrm{c}$ = Total activity	$\mathrm{R}\mathrm{E}=1-\left(\frac{\mathrm{N}\mathrm{R}}{\mathrm{N}\mathrm{a}\mathrm{c}}\right)$	Helleno et al. (2017)
10	Health level	$\mathrm{N}\mathrm{A}$ = Number of absent employees $\mathrm{N}\mathrm{E}$ = total number of employees	$\mathrm{H}\mathrm{E}=1-\left(\frac{\mathrm{N}\mathrm{A}}{\mathrm{N}\mathrm{E}}\right)$	Helleno et al. (2017); Paju et al. (2010)
11	Satisfaction level	$\mathrm{T}\mathrm{O}$ = number of employee turnover $\mathrm{N}\mathrm{E}$ = total number of employees	$\mathrm{S}\mathrm{E}=1-\left(\frac{\mathrm{T}\mathrm{O}}{\mathrm{N}\mathrm{E}}\right)$	Tett and Meyer (1993)

3.2. Relevant indicators selection via the Delphi method

Based on the TBL review literature indicators, the next step is the selection of industry-relevant indicators. Each indicator is assessed by industry experts using the Delphi method. The Delphi method is a group process for gathering expert opinions in a particular field (Yousuf, 2007). According to Noly (2018), the Delphi method allows experts to discuss and make decisions without having to meet. This method effectively solves problems based on expert opinion (Widiasih et al., 2015). The literature recommends 3 to 18 experts on the Delphi panel (Okoli & Pawlowski, 2004; Rowe & Wright, 1999). The Delphi method has been used for various issues in sustainability development, including the determination of indicator weights (Azevedo & Barros, 2017) and the green supply chain (Jiang et al., 2018).

In the Delphi method, the relevance of indicators to the company’s sustainability is determined by the average method (Feil et al., 2015). The cut-off of the relevant indicators is based on the analysis of weight average (WA) and level of consensus (LC; Feil et al., 2015; Miller, 2001). Hartini, Ciptomulyono, Anityasari et al. (2020) selected an indicator with WA > 4.0 and LC > 0.7. The formulas for WA and LC are presented in Equations 1(1) $\mathrm{W}\mathrm{A}=\frac{\sum \mathrm{S}{\mathrm{R}}_{\mathrm{i}}}{\mathrm{N}\mathrm{r}}$(1) and 2(2) $\mathrm{L}\mathrm{C}=\frac{\mathrm{F}\mathrm{N}\mathrm{R}}{\mathrm{N}\mathrm{r}}$(2) . Where $\mathrm{S}{\mathrm{R}}_{\mathrm{i}}$ shows the relevance assessment score of the $\mathrm{i}$-th respondent and $\mathrm{N}\mathrm{r}$ is the number of respondents. The number of respondents who stated they were relevant was notated as $\mathrm{F}\mathrm{N}\mathrm{R}$. The Delphi method uses five relevance rating scales from scores (1) Very irrelevant, (2). Irrelevant, (3). Neutral, (4). Relevant, and (5). Very relevant.

(1) $\mathrm{W}\mathrm{A}=\frac{\sum \mathrm{S}{\mathrm{R}}_{\mathrm{i}}}{\mathrm{N}\mathrm{r}}$(1)

(2) $\mathrm{L}\mathrm{C}=\frac{\mathrm{F}\mathrm{N}\mathrm{R}}{\mathrm{N}\mathrm{r}}$(2)

3.3. Interrelation indicator and indicator weighting

Multi-criteria decision-making is widely used in supply chain decision-making (Bag et al., 2021; A. K. Sahu et al., 2018; A. K., 2022), assessment logistics service (N. K. N. K. Sahu et al., 2015), and digital manufacturing barriers (Bag et al., 2022). In this research, multi-criteria decision-making is applied to asses interrelation indicators and indicator weighting. Dematel and ANP were proposed in this section to evaluate interrelation indicators and indicator weighting. Dematel is proposed to determine the relationship criterion. The ANP method is then used to determine the indicator TBL’s weighting. Using the Dematel method, the relationship between indicator can be evaluated. This procedure is based on Ranjbar et al. (2014) procedure. Using a scale of 0 (does not affect) to 4 (strongly affects), the Focus Group Discussion (FGD) team assesses the relationship between indicator. Results of the questionnaire are shown in a direct matrix ($\mathrm{B}$) as shown in Equation (3)(3) $\mathrm{B}=\left[\begin{array}{ccc}{\mathrm{b}}_{11}& {\mathrm{b}}_{1\mathrm{j}}& {\mathrm{b}}_{1\mathrm{n}}\\ {\mathrm{b}}_{\mathrm{i}1}& {\mathrm{b}}_{\mathrm{i}\mathrm{j}}& {\mathrm{b}}_{\mathrm{i}\mathrm{n}}\\ {\mathrm{b}}_{\mathrm{n}1}& {\mathrm{b}}_{\mathrm{n}\mathrm{j}}& {\mathrm{b}}_{\mathrm{n}\mathrm{n}}\end{array}\right]$(3) . As a result, it is normalized using Equations (4)(4) $\mathrm{s}=\frac{1}{\mathrm{m}\mathrm{a}\mathrm{x}{\sum }_{\mathrm{j}=1}^{\mathrm{n}}{\mathrm{b}}_{\mathrm{i}\mathrm{j}}}$(4) and (5(5) $\mathrm{X}=\mathrm{s}\mathrm{x}\mathrm{B}$(5) ), which produces a normalized matrix ($\mathrm{X}$) where $\mathrm{s}$ represents the normalization constant and $\mathrm{n}$ describes the number of indicator. The normalization results are also used to calculate the total relation matrix ($\mathrm{T}$) based on Equation (6)(6) $\mathrm{T}=\mathrm{X}{\left(\mathrm{I}-\mathrm{X}\right)}^{-1}$(6) . For the determination of $\mathrm{D}\mathrm{i}$ and $\mathrm{R}\mathrm{j}$, Equations (7)(7) $\mathrm{D}\mathrm{i}={\left[{\sum }_{\mathrm{j}=1}^{\mathrm{n}}{\mathrm{t}}_{\mathrm{i}\mathrm{j}}\right]}_{\mathrm{n}\mathrm{x}1}$(7) and (8(8) $\text{Rj}={\left[{\displaystyle {\sum }_{\text{i}=1}^{\text{n}}{\text{t}}_{\text{ij}}}\right]}_{1\text{xn}}^{}$(8) ) must be used in their entirety. The total T-row and total T-column are represented by $\mathrm{D}\mathrm{i}$ and $\mathrm{R}\mathrm{j}$, respectively. The prominence vector $\left(\mathrm{D}\mathrm{i}+\mathrm{R}\mathrm{j}\right)$ and the vector relation $\left(\mathrm{D}\mathrm{i}-\mathrm{R}\mathrm{j}\right)$ are obtained by using $\mathrm{D}\mathrm{i}$ and $\mathrm{R}\mathrm{j}$. The Impact Relation Map (IRM) and threshold value are determined at this point in Dematel’s procedure. In order to determine whether one set of indicator is related to another, we use a value called the threshold ($\alpha$). To determine whether the indicator are related, one must look at the $\mathrm{T}$ matrix’s value. The IRM, which serves as a network in the ANP procedure. Equation (9)(9) $T=\left[\begin{array}{ccc}{\mathrm{t}}_{11}& {\mathrm{t}}_{12}& {\mathrm{t}}_{13}\\ {\mathrm{t}}_{21}& {\mathrm{t}}_{22}& {\mathrm{t}}_{23}\\ {\mathrm{t}}_{31}& {\mathrm{t}}_{32}& {\mathrm{t}}_{33}\end{array}\right]$(9) shows an example of IRM on a T-matrices.

(3) $\mathrm{B}=\left[\begin{array}{ccc}{\mathrm{b}}_{11}& {\mathrm{b}}_{1\mathrm{j}}& {\mathrm{b}}_{1\mathrm{n}}\\ {\mathrm{b}}_{\mathrm{i}1}& {\mathrm{b}}_{\mathrm{i}\mathrm{j}}& {\mathrm{b}}_{\mathrm{i}\mathrm{n}}\\ {\mathrm{b}}_{\mathrm{n}1}& {\mathrm{b}}_{\mathrm{n}\mathrm{j}}& {\mathrm{b}}_{\mathrm{n}\mathrm{n}}\end{array}\right]$(3)

(4) $\mathrm{s}=\frac{1}{\mathrm{m}\mathrm{a}\mathrm{x}{\sum }_{\mathrm{j}=1}^{\mathrm{n}}{\mathrm{b}}_{\mathrm{i}\mathrm{j}}}$(4)

(5) $\mathrm{X}=\mathrm{s}\mathrm{x}\mathrm{B}$(5)

(6) $\mathrm{T}=\mathrm{X}{\left(\mathrm{I}-\mathrm{X}\right)}^{-1}$(6)

(7) $\mathrm{D}\mathrm{i}={\left[{\sum }_{\mathrm{j}=1}^{\mathrm{n}}{\mathrm{t}}_{\mathrm{i}\mathrm{j}}\right]}_{\mathrm{n}\mathrm{x}1}$(7)

(8) $\text{Rj}={\left[{\displaystyle {\sum }_{\text{i}=1}^{\text{n}}{\text{t}}_{\text{ij}}}\right]}_{1\text{xn}}^{}$(8)

(9) $T=\left[\begin{array}{ccc}{\mathrm{t}}_{11}& {\mathrm{t}}_{12}& {\mathrm{t}}_{13}\\ {\mathrm{t}}_{21}& {\mathrm{t}}_{22}& {\mathrm{t}}_{23}\\ {\mathrm{t}}_{31}& {\mathrm{t}}_{32}& {\mathrm{t}}_{33}\end{array}\right]$(9)

In this research, the ANP procedure is used to calculate a weighting indicator based on interrelation. The ANP method is based on Saaty (Saaty, 2004) and Yang et al. (2008). Pairwise comparisons on a scale of 1 (equal importance) to 9 (absolutely more important) are used to evaluate IRM results on Dematel. Equation (10)(10) $\mathrm{A}=\begin{array}{c}\mathrm{C}1\\ \mathrm{C}2\\ \mathrm{C}\mathrm{n}\end{array}\left[\begin{array}{ccc}{\mathrm{a}}_{11}& {\mathrm{a}}_{12}& \cdots \hspace{0.17em}\\ {\mathrm{a}}_{12}& {\mathrm{a}}_{22}& \cdots \hspace{0.17em}\\ {\mathrm{a}}_{\mathrm{n}1}& {\mathrm{a}}_{\mathrm{n}2}& \cdots \hspace{0.17em}\end{array}\begin{array}{c}{\mathrm{a}}_{1\mathrm{n}}\\ {\mathrm{a}}_{2\mathrm{n}}\\ {\mathrm{a}}_{\mathrm{n}\mathrm{n}}\end{array}\right]$(10) displays formula the unweighted supermatrix from the pairwise comparison assessment indicator ($\mathrm{A}$). In addition, the weighted supermatrix is calculated in the next step. Transforming the initial $\mathrm{T}$ matrix to the a new matrix α-cut total relation matrix ($\mathrm{T}\alpha$) is an important step in the analysis process. Equation (11)(11) $\mathrm{T}\alpha =\left[\begin{array}{ccc}{\mathrm{t}}_{11}^{\propto }& {\mathrm{t}}_{1\mathrm{j}}^{\propto }& {\mathrm{t}}_{1\mathrm{n}}^{\propto }\\ {\mathrm{t}}_{\mathrm{i}1}^{\propto }& {\mathrm{t}}_{\mathrm{i}\mathrm{j}}^{\propto }& {\mathrm{t}}_{\mathrm{i}\mathrm{n}}^{\propto }\\ {\mathrm{t}}_{\mathrm{n}1}^{\propto }& {\mathrm{t}}_{\mathrm{n}\mathrm{j}}^{\propto }& {\mathrm{t}}_{\mathrm{n}\mathrm{n}}^{\propto }\end{array}\right]$(11) shows the Tα matrix, which is created by assigning a value of 0 to any matrix value less than the threshold value. It equations are used to calculate the weighted supermatrix ($\mathrm{A}\mathrm{w}$) in Equations (12)(12) ${\mathrm{d}}_{\mathrm{i}}={\sum }_{\mathrm{j}=1}^{\mathrm{n}}{\mathrm{t}}_{\mathrm{i}\mathrm{j}}^{\propto }$(12) , (13(13) $\mathrm{T}\mathrm{s}=\left[\begin{array}{ccc}{\mathrm{t}}_{11}^{\mathrm{s}}& {\mathrm{t}}_{1\mathrm{j}}^{\mathrm{s}}& {\mathrm{t}}_{1\mathrm{n}}^{\mathrm{s}}\\ {\mathrm{t}}_{\mathrm{i}1}^{\mathrm{s}}& {\mathrm{t}}_{\mathrm{i}\mathrm{j}}^{\mathrm{s}}& {\mathrm{t}}_{\mathrm{i}\mathrm{n}}^{\mathrm{s}}\\ {\mathrm{t}}_{\mathrm{n}1}^{\mathrm{s}}& {\mathrm{t}}_{\mathrm{n}\mathrm{j}}^{\mathrm{s}}& {\mathrm{t}}_{\mathrm{n}\mathrm{n}}^{\mathrm{s}}\end{array}\right]$(13) ), and (14(14) $\mathrm{A}\mathrm{W}=\left[\begin{array}{ccc}{\mathrm{t}}_{11}^{\mathrm{s}}\mathrm{x}{\mathrm{a}}_{11}& {\mathrm{t}}_{1\mathrm{j}}^{\mathrm{s}}\mathrm{x}{\mathrm{a}}_{12}& \cdots \hspace{0.17em}\\ {\mathrm{t}}_{\mathrm{i}1}^{\mathrm{s}}\mathrm{x}{\mathrm{a}}_{12}& {\mathrm{t}}_{\mathrm{i}\mathrm{j}}^{\mathrm{s}}\mathrm{x}{\mathrm{a}}_{22}& \cdots \hspace{0.17em}\\ {\mathrm{t}}_{\mathrm{n}1}^{\mathrm{s}}\mathrm{x}{\mathrm{a}}_{\mathrm{n}1}& {\mathrm{t}}_{\mathrm{n}\mathrm{j}}^{\mathrm{s}}\mathrm{x}{\mathrm{a}}_{\mathrm{n}2}& \cdots \hspace{0.17em}\end{array}\begin{array}{c}{\mathrm{t}}_{1\mathrm{n}}^{\mathrm{s}}\mathrm{x}{\mathrm{a}}_{1\mathrm{n}}\\ {\mathrm{t}}_{\mathrm{i}\mathrm{n}}^{\mathrm{s}}\mathrm{x}{\mathrm{a}}_{2\mathrm{n}}\\ {\mathrm{t}}_{\mathrm{n}\mathrm{n}}^{\mathrm{s}}\mathrm{x}{\mathrm{a}}_{\mathrm{n}\mathrm{n}}\end{array}\right]$(14) ). Where, $\mathrm{d}\mathrm{i}$ is the number of rows in the matrix T. The calculation of the Limiting supermatrix, which is formulated in Equation (15)(15) $\underset{\mathrm{k}\to \mathrm{\infty }}{lim}{\mathrm{W}}_{\mathrm{w}}^{\mathrm{k}}$(15) , is the final step in the ANP procedure. The weight of each TBL indicator is the result of the Limiting supermatrix.

$\begin{array}{ccc}\mathrm{C}1& \mathrm{C}2& \cdots \end{array}\mathrm{C}\mathrm{n}$

(10) $\mathrm{A}=\begin{array}{c}\mathrm{C}1\\ \mathrm{C}2\\ \mathrm{C}\mathrm{n}\end{array}\left[\begin{array}{ccc}{\mathrm{a}}_{11}& {\mathrm{a}}_{12}& \cdots \hspace{0.17em}\\ {\mathrm{a}}_{12}& {\mathrm{a}}_{22}& \cdots \hspace{0.17em}\\ {\mathrm{a}}_{\mathrm{n}1}& {\mathrm{a}}_{\mathrm{n}2}& \cdots \hspace{0.17em}\end{array}\begin{array}{c}{\mathrm{a}}_{1\mathrm{n}}\\ {\mathrm{a}}_{2\mathrm{n}}\\ {\mathrm{a}}_{\mathrm{n}\mathrm{n}}\end{array}\right]$(10)

(11) $\mathrm{T}\alpha =\left[\begin{array}{ccc}{\mathrm{t}}_{11}^{\propto }& {\mathrm{t}}_{1\mathrm{j}}^{\propto }& {\mathrm{t}}_{1\mathrm{n}}^{\propto }\\ {\mathrm{t}}_{\mathrm{i}1}^{\propto }& {\mathrm{t}}_{\mathrm{i}\mathrm{j}}^{\propto }& {\mathrm{t}}_{\mathrm{i}\mathrm{n}}^{\propto }\\ {\mathrm{t}}_{\mathrm{n}1}^{\propto }& {\mathrm{t}}_{\mathrm{n}\mathrm{j}}^{\propto }& {\mathrm{t}}_{\mathrm{n}\mathrm{n}}^{\propto }\end{array}\right]$(11)

(12) ${\mathrm{d}}_{\mathrm{i}}={\sum }_{\mathrm{j}=1}^{\mathrm{n}}{\mathrm{t}}_{\mathrm{i}\mathrm{j}}^{\propto }$(12)

$\mathrm{T}\mathrm{s}=\left[\begin{array}{ccc}{\mathrm{t}}_{11}^{\propto }/\mathrm{d}1& {\mathrm{t}}_{1\mathrm{j}}^{\propto }/\mathrm{d}1& {\mathrm{t}}_{1\mathrm{n}}^{\propto }/\mathrm{d}1\\ {\mathrm{t}}_{\mathrm{i}1}^{\propto }/\mathrm{d}\mathrm{i}& {\mathrm{t}}_{\mathrm{i}\mathrm{j}}^{\propto }/\mathrm{d}\mathrm{i}& {\mathrm{t}}_{\mathrm{i}\mathrm{n}}^{\propto }/\mathrm{d}\mathrm{i}\\ {\mathrm{t}}_{\mathrm{n}1}^{\propto }/\mathrm{d}3& {\mathrm{t}}_{\mathrm{n}\mathrm{j}}^{\propto }/\mathrm{d}3& {\mathrm{t}}_{\mathrm{n}\mathrm{n}}^{\propto }/\mathrm{d}3\end{array}\right]$

(13) $\mathrm{T}\mathrm{s}=\left[\begin{array}{ccc}{\mathrm{t}}_{11}^{\mathrm{s}}& {\mathrm{t}}_{1\mathrm{j}}^{\mathrm{s}}& {\mathrm{t}}_{1\mathrm{n}}^{\mathrm{s}}\\ {\mathrm{t}}_{\mathrm{i}1}^{\mathrm{s}}& {\mathrm{t}}_{\mathrm{i}\mathrm{j}}^{\mathrm{s}}& {\mathrm{t}}_{\mathrm{i}\mathrm{n}}^{\mathrm{s}}\\ {\mathrm{t}}_{\mathrm{n}1}^{\mathrm{s}}& {\mathrm{t}}_{\mathrm{n}\mathrm{j}}^{\mathrm{s}}& {\mathrm{t}}_{\mathrm{n}\mathrm{n}}^{\mathrm{s}}\end{array}\right]$(13)

(14) $\mathrm{A}\mathrm{W}=\left[\begin{array}{ccc}{\mathrm{t}}_{11}^{\mathrm{s}}\mathrm{x}{\mathrm{a}}_{11}& {\mathrm{t}}_{1\mathrm{j}}^{\mathrm{s}}\mathrm{x}{\mathrm{a}}_{12}& \cdots \hspace{0.17em}\\ {\mathrm{t}}_{\mathrm{i}1}^{\mathrm{s}}\mathrm{x}{\mathrm{a}}_{12}& {\mathrm{t}}_{\mathrm{i}\mathrm{j}}^{\mathrm{s}}\mathrm{x}{\mathrm{a}}_{22}& \cdots \hspace{0.17em}\\ {\mathrm{t}}_{\mathrm{n}1}^{\mathrm{s}}\mathrm{x}{\mathrm{a}}_{\mathrm{n}1}& {\mathrm{t}}_{\mathrm{n}\mathrm{j}}^{\mathrm{s}}\mathrm{x}{\mathrm{a}}_{\mathrm{n}2}& \cdots \hspace{0.17em}\end{array}\begin{array}{c}{\mathrm{t}}_{1\mathrm{n}}^{\mathrm{s}}\mathrm{x}{\mathrm{a}}_{1\mathrm{n}}\\ {\mathrm{t}}_{\mathrm{i}\mathrm{n}}^{\mathrm{s}}\mathrm{x}{\mathrm{a}}_{2\mathrm{n}}\\ {\mathrm{t}}_{\mathrm{n}\mathrm{n}}^{\mathrm{s}}\mathrm{x}{\mathrm{a}}_{\mathrm{n}\mathrm{n}}\end{array}\right]$(14)

(15) $\underset{\mathrm{k}\to \mathrm{\infty }}{lim}{\mathrm{W}}_{\mathrm{w}}^{\mathrm{k}}$(15)

3.4. Indicator efficiency assessment and sustainable VSM mapping

This section describes the mapping of indicators using Sustainable VSM. The Sustainable VSM is a classic VSM combined with a TBL indicator. It is expected to be able to describe the company’s performance using economic, environmental, and social aspects (Faulkner & Badurdeen, 2014). As described in the previous section, the relevant indicators were selected based on the Delphi method. Furthermore, the selected indicators are calculated for efficiency based on the formula presented in Table 3. The efficiency values for each indicator are used for mapping sustainable VSM. The efficiency values for each indicator in sustainable VSM are classified using the Traffic Light System (TLS) principle. Three colors are used for indicator mapping, namely red with a percentage of indicator value < 60%, indicating the indicator score is still below the target. The yellow color represents indicators with a 60%-90% indicating the achieved score needs improvement. The green color represents the percentage of indicators > 90%, indicating an excellent indicator (Hartini, Ciptomulyono, Anityasari et al., 2020). Furthermore, the efficiency value of each indicator is used to calculate the Manufacturing Sustainability Score (MSS), which is described in the following subsection.

3.5. Score manufacturing sustainability assessment

In this section, the MSS is based on the value of the efficiency of the indicator and the weight of the indicator. The MSS value is generated by multiplying the efficiency of the indicator with the weight of the indicator (Huang & Badurdeen, 2018; De Silva et al., 2009). The MSS calculation formula can be seen in equation (16)(16) $\mathrm{M}\mathrm{S}\mathrm{S}={\sum }_{\mathrm{i}}^{\mathrm{n}}\mathrm{W}\mathrm{i}.\mathrm{E}\mathrm{i}$(16) . Where, $\mathrm{W}\mathrm{i}$ shows the weight of the $\mathrm{i}$-indicator generated by the ANP method, and $\mathrm{E}\mathrm{i}$ shows the $\mathrm{i}$-indicator efficiency score. The number of relevant indicators is denoted as $\mathrm{n}$.

(16) $\mathrm{M}\mathrm{S}\mathrm{S}={\sum }_{\mathrm{i}}^{\mathrm{n}}\mathrm{W}\mathrm{i}.\mathrm{E}\mathrm{i}$(16)

MSS scores are also classified using the TLS principle, which is also used to classify efficiency indicators. The MSS percentage of 60% is shown in red, indicating that the score is still below the target. The percentage of 60% – 90% indicates that the indicator needs to be improved (which is depicted in yellow). Furthermore, an MSS percentage greater than 90% indicates that the indicator is on target (depicted in green).

4. Case study application

A case study was conducted in the wood furniture industry in Indonesia, which is of the make-to-order type. This study involved ten experts in assessing the relevant TBL indicators. Each expert assessed the relevance score of the indicators used. Furthermore, the Delphi method uses the assessment score data to select relevant indicators according to the problems in the company. The results of the assessment of the relevance of each indicator are shown in Table 4. It shows two irrelevant indicators, namely the inventory indicator and the material consumption indicator. These results indicated that nine relevant indicators are used to assess manufacturing sustainability.

Table 4. Results of the assessment of the relevance of each indicator.

Factors	Indicators	Relevance					WA	LC
		1	2	3	4	5
Economic	Time				4	6	4.6	1
	Inventory		3	4	2	1	3.1	0.3
	Quality				3	7	4.7	1
	Cost					10	5	1
Enviroment	Material		3	3	2	2	3.3	0.4
	Energy			2	5	3	4.1	0.8
	Waste recycle				1	9	4.9	1
Social	Satisfaction level					10	5	1
	Health level					10	5	1
	Safety level				1	9	4.9	1
	Employee training			2	3	5	4.3	0.8

The experts also conducted a Focus Group Discussion (FGD) to assess the relationship between the indicators. The results of the assessment of the relationship between indicators are presented in Table 5. It used the Dematel method to assess the relationship between indicators of the TBL variables. The results of the α-cut total relation matrix are shown in Table 6 with a threshold value of 0.430. It shows that time and quality indicators are the most dispatcher indicators, followed by cost and energy indicators. In addition, the safety level is the most receiver indicator influenced by other indicators, followed by the health level. The α-cut total relation matrix is also used to describe the network in ANP. The ANP network in this problem is depicted in Figure 2. This study proposes three main clusters of the ANP network. Cluster 1 shows the comparison between indicators and aspects of TBL. Next, a comparison of each aspect of TBL is presented in cluster 2. Finally, cluster 3 is a pairwise comparison between TBL indicators. Super decision software was used as a tool to carry out ANP procedures.

Figure 2. Network ANP.

Table 5. Assessment of the relationship between indicators.

Indicators	Time	Quality	Cost	Energy	Waste recycle	Satisfaction level	Health level	Safety level	Employee training
Time	0	4	4	3	2	1	1	1	3
Quality	4	0	3	3	4	4	0	0	1
Cost	2	3	0	2	1	3	1	1	2
Energy	3	3	3	0	1	1	0	0	0
Waste recycle	3	4	4	2	0	0	3	1	3
Satisfaction level	4	3	3	2	3	0	2	3	2
Health level	4	4	4	2	2	4	0	3	1
Safety level	3	3	3	2	3	3	3	0	3
Employee training	4	4	3	3	4	1	1	4	0

Table 6. Results of α -cut total relation matrix.

Indicators	Time	Quality	Cost	Energy	Waste recycle	Satisfaction level	Health level	Safety level	Employee training
Time	0.440	0.607	0.593	0.451	0.000	0.000	0.000	0.000	0.000
Quality	0.581	0.460	0.560	0.446	0.477	0.000	0.000	0.000	0.000
Cost	0.449	0.497	0.000	0.000	0.000	0.000	0.000	0.000	0.000
Energy	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000
Waste recycle	0.588	0.648	0.633	0.443	0.000	0.000	0.000	0.000	0.000
Satisfaction level	0.675	0.669	0.654	0.482	0.519	0.000	0.000	0.000	0.000
Health level	0.717	0.743	0.728	0.512	0.515	0.533	0.000	0.000	0.000
Safety level	0.675	0.701	0.685	0.504	0.544	0.484	0.000	0.000	0.450
Employee training	0.707	0.737	0.687	0.543	0.580	0.000	0.000	0.000	0.000

FGD data for pairwise comparison assessment between indicators and aspects of TBL is presented in Table A1-A12 in Appendix A. Data for each indicator is collected from 6 work stations, including measurement, cutting, planer, router trimmer, assembly, and finishing. The components of the data collected on each indicator are presented in Table 3. Furthermore, these are used for calculating efficiency and manufacturing sustainability scores.

5. Results and discussion

5.1. Assessment results

The results of the weighting of indicators using ANP are presented in Table 7. The index value of the consistency ratio of the pairwise comparisons is below 10%, which indicates that the data is consistent. This study shows that quality has a high weight, followed by time and cost indicators. Quality is the most critical indicator of manufacturing sustainability because it satisfies customers. This study’s results align with research conducted by Hartini, Ciptomulyono, Anityasari et al. (2020). Based on the research by Saidani and Arifin (2012), product quality can influence consumer satisfaction and buying interest. Therefore, the company must make a product quality plan oriented to product quality, which will be an advantage that can be used to face the competition (Iswanto et al., 2013).

Table 7. Weight indicators based on ANP.

Factors	Indicators	Weight
Economic	Time	0.260
	Quality	0.285
	Cost	0.130
Enviroment	Energy	0.104
	Waste recycling	0.150
Social	Satisfaction level	0.010
	Health level	0.023
	Safety level	0.023
	Employee training	0.015

In addition, the time indicator is the most important because the waste of time that is not added value can reduce the company’s performance. This indicator also has the effect of decreasing the performance of the cost indicator. Cost also has a critical role in achieving sustainable development. Cost is the equivalent value sacrificed to get a product that is expected to be helpful in the present and future. The costs in the production process include the cost of raw materials, labor costs, and factory overhead (Vinodh et al., 2014). The cost indicator in the company is used as a measure of the success of a business.

The results of the calculation of the efficiency of each indicator and MSS are shown in Table 8. Furthermore, the mapping of sustainable VSM is depicted in Figure 3. These results show that the time, quality, cost, and energy indicators contribute to the highest indicator score. Based on the Sus-VSM and TLS mapping results, eight indicators have moderate (yellow) and low (red) efficiency values. In the time indicator, three work stations have low-efficiency values: cutting (43.5%), router trimmer (53.9%), and finishing (43.8%). The trimmer router workstation has the lowest quality indicator efficiency (0.7%). The waste recycling and employee training indicators have an efficiency value of 0%. The health level indicator contributes 46% efficiency, and the Safety level indicator has 50% efficiency. The MSS in this industry is 66.97%, which shows that the performance of manufacturing sustainability in the production line is still lacking.

Figure 3. Mapping of SustainableVSM.

Table 8. Calculation Results of Indicator and Manufacturing Sustainability Score.

Factors	Indicators	Efficiency (%)	Weight	Score indicators	MSS
Economic	Time	63.6	0.260	16.56	66.97%
	Quality	85.2	0.285	24.31
	Cost	99.3	0.130	12.87
Enviroment	Energy	100	0.104	10.37
	Waste recycle	0	0.150	0.00
Social	Satisfaction level	62	0.010	0.62
	Health level	46	0.023	1.07
	Safety level	50	0.023	1.17
	Employee training	0	0.015	0.00

5.2. Practical and managerial implications

This study contains several managerial insights that are not only academic but also managerial. The implications of each point of view are clearly explained below.

5.2.1. Academic implications

Manufacturing companies can better assess, map, and improve operations using this method to achieve manufacturing sustainability. This proposed method is intended to fill gaps in previous research that has extensively used the TBL concept but falls short in analyzing the interaction between indicators. Using this method, the decision-maker can explain the significance of the interrelation of each indicator TBL. The production line is a manufacturing system component, contributing to economic, social, and environmental issues. With this understanding, it is clear that the environment, economy, and society all play important roles in long-term development and interact with one another. The interaction of indicators influences the weight of importance, influencing the manufacturing sustainability score.

According to the study, inefficient production lines in terms of TBL impact the manufacturing sustainability score. As a result of this research, academic research in other industrial sectors needs further improvement and investigation. Previous studies on sustainable manufacturing practices have not been thoroughly investigated, but they are a promising beginning.

5.2.2. Managerial implications and recommendations

This study assists industry managers in focusing on sustainable manufacturing practices to increase their chances of implementing sustainable manufacturing. Managers are encouraged to make the best decisions possible to improve the manufacturing sustainability score in today’s increasingly complex business environment.

According to the research findings, six indicators have low-efficiency values: waste recycling, employee training, health level, safety level, time, and quality. On the other hand, the energy indicator has high efficiency in this industry (Utama, 2022; Utama & Widodo, 2021). It demonstrates that the company has implemented an efficient manufacturing process. Scheduling correctly has improved energy efficiency in the manufacturing process (Utama, Widodo, Ibrahim, Hidayat et al., 2020; Utama et al., 2019). The following are recommendations for improving performance in each indicator:

1. Recommendations for improving waste recycling indicators’ performance include collaborating with third parties for waste management. For example, the industry can use wood waste to make fertilizers and handicrafts.

2. Recommendations for improving employee training indicators include training to increase worker skills in making a product in particular and in the use of machines. Defects in a part or unit due to a lack of skills can affect production time and costs.

3. The proposed improvement for the health level indicator is to apply work discipline and follow up on wood assault waste to minimize employee health problems.

4. Proposed enhancements to safety level indicators, namely increased worker discipline in using personal protective equipment. Many workers who do not use personal protective equipment are discovered, causing the impact felt by workers and businesses. The company may incur losses in terms of both time and money.

5. Proposed value-add improvements, such as eliminating non-value-added activities such as waiting. Furthermore, businesses must implement 5S, inspect each workstation, and develop standard operational procedures.

6. Finally, the quality indicator recommendation is to conduct inspections at each work station to detect defects before the product reaches the consumer.

6. Conclusion

The primary goal of this research is to develop a new Manufacturing Sustainability Performance Assessment (MSPA) framework for the manufacturing process line. In the manufacturing process line, a new framework with a hybrid method has been successfully developed for MSPA. The Delphi method was used to choose relevant indicators. Relevant indicators are calculated for efficiency and mapped into sustainable VSM and TLS. The MSS calculation is based on the weight of the Dematel-ANP method indicators and the efficiency of each indicator. This proposed framework is also used in the Indonesian furniture industry. The case study results show that the furniture industry’s manufacturing sustainability performance needs improvement (yellow). Furthermore, it demonstrates that the proposed method can be used to solve real-world problems.

This study makes two significant contributions to the MSPA. First, as a novel contribution, we propose a new MSPA framework for use in sustainable manufacturing firms. The proposed framework incorporates Delphi-Dematel-ANP and long-term VSM procedures. The second contribution is that we empirically tested the proposed framework in industrial furniture manufacturing companies. It is done because no MSPA assessment includes interrelationship indicators.

As with previous studies, the limitations of the research presented in this paragraph also apply to this study. First, our study focuses on manufacturing companies that produce a product variant, whereas the current MSPA assessment focuses on a single product. Second, this study is based on data collected over a specific time horizon and with deterministic data types. This problem assumes that the information obtained is clear. Third, we have not included worker mental and physical workload indicators that were not considered in the MSPA. Our framework will be complete if we include workers’ mental and physical workload indicators. Fourth, this case study only assesses the feasibility of the proposed model for evaluating MSPA, particularly in the furniture industry. It should be noted that the indicators used in this study are only relevant for evaluating industries that are part of the furniture sector. Finally, the current framework should be investigated further under different conditions, which may provide different insights into the current framework. The following is a description of the future research direction:

• More research is needed to consider multiple product variants in the MSPA assessment so that the new framework can evaluate manufacturing firms producing multiple products.

• In order to represent uncertainty, MSPA must consider fuzzy and stochastic data types.

• In order to fully assess manufacturing’s sustainability performance, additional research must consider worker mental and physical workload indicators.

• Additional research on various manufacturing industries is required to validate the model and improve the indicators in the proposed framework.

Data availability

All data generated or analyzed during this study are included in this article.

Acknowledgments

The researcher would like to thank the University of Muhammadiyah Malang for supporting the research. In addition, the researcher would also like to extend their gratitude to the Department of Industrial Engineering Optimization Laboratory for the help of facilities.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Appendix A

Table A1. Pairwise comparisons between TBL factors.

	Economic	Enviroment	Social
Economic	1	2	3
Enviroment	1/2	1	2
Social	1/3	1/2	1

Table A2. Pairwise comparison of indicators on economic factors.

	Cost	Quality	Time
Cost	1	1/2	1/2
Quality	2	1	1
Time	2	1	1

Table A3. Pairwise comparison of indicators on environmental factors.

	Energy	Waste
Energy	1	1/2
Waste	2	1

Table A4. Pairwise comparison of indicators on social factors.

	Employee	Health	Safety	Satisfaction
Employee	1	1/2	1/2	2
Health	2	1	1	3
Safety	2	1	1	3
Satisfaction	1/2	1/3	1/3	1

Table A5. Pairwise comparison of indicators that affect cost indicator.

	Quality	Time
Quality	1	2
Time	1/2	1

Table A6. Pairwise comparison of indicators that affect Employee training indicators.

	Cost	Energy	Quality	Time	Waste
Cost	1	1	1/2	1/2	2
Energy	1	1	1/2	1/2	2
Quality	2	2	1	1	2
Time	2	2	1	1	2
Waste	1/2	1/2	1/2	1/2	1

Table A7. Pairwise comparison of indicators that affect the indicator Health.

	Cost	Energy	Quality	Satisfaction	Time	Waste
Cost	1	2	1/2	1	1	2
Energy	1/2	1	1/2	1/2	1/3	3
Quality	2	2	1	2	2	2
Satisfaction	1	2	1/2	1	1	2
Time	1	3	1/2	1	1	2
Waste	1/2	1/3	1/2	1/2	1/2	1

Table A8. Pairwise comparison of indicators that affect Quality indicators.

	Cost	Energy	Quality	Time	Waste
Cost	1	2	1/2	1/2	2
Energy	1/2	1	1/3	1/3	1
Quality	2	3	1	1	2
Time	2	3	1	1	2
Waste	1/2	1	1/2	1/2	1

Table A9. Pairwise comparison of indicators that affect safety indicators.

	Cost	Employee	Energy	Quality	Satisfaction	Time
Cost	1	1/2	1/2	1/3	2	1
Employee	2	1	1	1/2	2	2
Energy	2	1	1	1/2	2	2
Quality	3	2	2	1	4	2
Satisfaction	1/2	1/2	1/2	1/4	1	1/2
Time	1	1/2	1/2	1/2	2	1

Table A10. Pairwise comparison of indicators that affect the Satisfaction indicator.

	Cost	Energy	Quality	Time	Waste
Cost	1	1/2	1/2	1/2	2
Energy	2	1	1	1	2
Quality	2	1	1	1	2
Time	2	1	1	1	2
Waste	1/2	1/2	1/2	1/2	1

Table A11. Pairwise comparison of indicators that affect Time indicators.

	Cost	Energy	Quality	Time
Cost	1	3	1/2	1/2
Energy	1/3	1	1/4	1/4
Quality	2	4	1	1
Time	2	4	1	1

Table A12. Pairwise comparison of indicators that affect Waste Recycle indicators.

	Cost	Energy	Quality	Time
Cost	1	2	1/2	1/2
Energy	1/2	1	1/3	1/3
Quality	2	3	1	1
Time	2	3	1	1