A Model for a Green, Lean Sustainable Improvement with Performance Measurement

Abstract

With a turbulent market and a desire to withstand force, many companies lower the production price of their products using different approaches. On the one hand, they often use lean principles to reduce production costs, while they are compelled by the government to adopt green practices on the other. The proposed approach (model) is a combination of green, lean sustainable improvement, leading a company through the green, lean process with measurable results and optimization functions. In addition to its return on investment (ROI), a company can calculate its output (from material to emission flow) based on its input. Based on the radar chart we have produced, they can also know what they have achieved and in which direction they need to optimize to develop in as green and sustainable a way as possible. One example of a Slovenian automotive company demonstrates how the implementation of the proposed measures can effectively reduce environmental impact, while also ensuring a quick return on investment.

1. Introduction and Literature Overview

Many barriers in green, lean implementation can be found in the literature. Anass Cherrafi [1] identified 15 common barriers to implementing green and lean practices, including a lack of environmental awareness, a fear of failure, poor quality of human resources, a lack of expertise training and education, a lack of statistical green and lean thinking, inappropriate identification of areas and activities to be made lean and green, a lack of Kaizen culture, a lack of visual and statistical control, a lack of government support, high cost, a lack of communication and cooperation between departments, a lack of top management involvement, resistance to change, and a poor culture of cooperation, separating environmental and continuous improvement decisions. Cherrafi also pointed out key commonalities between lean and green implementations, such as waste reduction, emphasis on people and the organization, lead time reduction, supply chain relationships, service-level KPIs, and common tools and processes.

Zhu and Zhang [2] categorized 17 of the identified barriers into four groups. The first group includes barriers to green, lean practices, while the second group addresses barriers to the adoption and implementation of lean, green strategies. The third group addresses barriers related to technological advancement and management, and the fourth group includes barriers related to knowledge, communication, and conflict.

Wei Dong Leong et al. [3] examined the application and impact of lean and green approaches in the manufacturing industry and highlighted the similarities and benefits of both strategies. They also pointed out a significant lack of political support and government commitment to implementing lean and green practices. Implementing green practices can have a positive impact on a company’s performance. It means that a company adopts proactive environmental behaviors, i.e., actions aimed at reducing pollution and improving the company’s economic performance [4]. Srivastava [5] focuses on supply chain management and claims that the adoption of environmentally friendly practices can reduce the negative environmental impact of industrial operations while maintaining the desired levels of quality, cost, reliability, and energy efficiency. Carvalhoa [6] noted that not all companies within the same supply chain can be completely lean or green; instead, trade-offs are required within individual companies to meet both the environmental and economic constraints of the supply chain. They also proposed a strategic framework to support the design of eco-efficient supply chains.

Inman [7] used survey data to confirm that lean manufacturing practices are positively correlated with environmental and operational performance, and that green supply chain management practices are positively associated with environmental performance. However, no direct correlation was found between green supply chain practices and operational performance. Although lean practices have a direct impact on environmental performance, their indirect impact is even stronger when complemented by green practices, highlighting their synergy.

A combination of lean and green approaches aims at maximizing the efficiency of organizations and their supply chains [8]. According to [9], studies of organizations that are both lean and green suggest that although the two approaches have different drivers and methodologies, they are compatible and can even create synergy. Bhattacharya [10] also found that lean and green practices share common characteristics, and their integration improves sustainability performance in the economic, environmental, and social dimensions of organizational production systems.

Fercoq et al. [11] investigated the individual contributions of lean and green methods of reducing solid waste in manufacturing. They demonstrated that combining the concepts of reduction, reuse, and recovery with a focus on waste elimination (Muda), as in lean management, increases the effectiveness of waste minimization programs in manufacturing.

Singh et al. [12] categorized and analyzed twelve barriers to green, lean implementation using a cause-and-effect approach. They identified “resistance to change”, a “lack of top management commitment”, and a “lack of employee training” as the most important barriers based on the results of the DEMATEL approach.

Green Lean Six Sigma (GLS) was recognized by [13] as a comprehensive approach that reduces negative environmental impacts while delivering high-quality products. To meet customer requirements focused on sustainability, companies need to understand the key elements and methods of implementing the GLS approach.

The concept of a sustainable lean model was introduced by [14]. This concept evolves from lean manufacturing (focusing on rapid, resilient responses to flow, variation, and disruption) to lean enterprise (designing organizations that drive change and operational excellence) and, finally, to sustainable lean practice or Lean 5.0, which focuses on cultivating a culture that sustains change and excellence. Sawhney et al. [14] also developed a sustainable lean model with three phases: system prognosis (defining system-related problems and aligning continuous improvement with organizational goals), system diagnosis (improving system throughput by addressing disruption and variability), and sustaining system health (addressing cultural differences, identifying employee resistance, etc.).

The Business Overall Performance and Sustainability Effectiveness (BOPSE) indicator was designed by [15]. BOPSE is equivalent to ((Sustainability + OEE)/2), where sustainability consists of a product of economic, environmental, and social percentages, and the overall equipment effectiveness (OEE) of availability, performance, and quality percentages.

2. Tropical Algebra

In the proposed model that will be introduced in the next section, we will formulate optimization problems for two target functions, which will also be expressed in tropical algebra terminology. Tropical algebra is a term that unifies isomorphic algebraic mathematical systems (max-plus algebra, min-plus algebra, max-times algebra, and min-times algebra) and provides (together with its geometric counterpart—tropical geometry) an attractive way of describing a class of non-linear problems appearing, for instance, in the scheduling of manufacturing and transportation networks, discrete time event-dynamic systems and their optimal control, combinatorial optimization, mathematical physics, information technology, system theory of computer networks, DNA analysis (the study of phylogenetic trees), machine learning and deep neural networks, medicine (the non-local diffusion of networks arising from medical data), differential equations with delay (the asymptotic study of slowly oscillating solutions of a class of differential equations with delay that arise in models from biology, social sciences, crystal growth, etc.), and graph theory, etc. See, e.g., [16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33], and the references cited therein. The usefulness of tropical algebra arises from the fact that these non-linear problems become linear when described in the language of tropical algebra. Its methods also turned out to be very useful for solving certain problems in linear algebra and graph theory—see, e.g., [25,30,31,32,33] and the references cited therein.

Max-plus algebra is the (semialgebra) algebra over the ordered, idempotent semiring (in fact, semifield) R_max = R ∪ {−∞} (where R denotes a set of real numbers) equipped with the operations of the addition a ⊕ b = max {a,b} and the multiplication a ⊙ b = a + b, with the unit elements −∞ (for max-plus addition ⊕) and 0 (for max-plus multiplication ⊙). Here, we define −∞ ⊙ a = a ⊙ −∞ = −∞. The max-plus power is defined by a ⊙ k = ka (the standard multiplication) for the real numbers a and k. This notation is used to write ${\oplus }_{i=1}^k{a}_i=\mathit{max}\{a_1,\dots ,a_k\}$ and ${\odot }_{i=1}^k{a}_i={\sum }_{i=1}^k{a}_i$ for $a_1,\dots ,a_k\in R_{max}$. As in standard arithmetic, the operations of addition and multiplication are associative and commutative, and multiplication is distributive over addition. Matrix and polynomial operations are defined similarly to their standard counterparts, with the max-plus operations replacing the standard operations.

Max-plus algebra is isomorphic to min-plus algebra, which is the semifield ${ R}_{min}=R\cup \{\infty \}$, equipped with the operations of addition $a{\oplus }^{\prime } b=min\{a,b\}$ and multiplication a ${\odot }^{\prime } b=a+b$, with the unit elements ∞ (for min plus addition ⊕’) and 0 (for max-plus multiplication ⊙’). Here, we define $\infty {\odot }^{\prime }a=a {\odot }^{\prime } \infty =\infty$. We also write ${{\oplus }^{\prime } }_{i=1}^k{a}_i=min \{a_1,\dots ,a_k\}$ and ${{\odot }^{\prime } }_{i=1}^k{a}_i={\sum }_{i=1}^k{a}_i$ for $a_1,\dots ,a_k\in R_{min}$. Thus, for ${ a}_1,\dots ,a_k\in R$, we have ${\odot }_{i=1}^k{a}_i={{\oplus }^{\prime } }_{i=1}^k{a}_i={\sum }_{i=1}^k{a}_i.$

3. Proposed Model

Based on our experience of lean practices and the literature review on green, lean practices, a model for green, lean sustainable production was built on the basis of the company’s culture, which has to be prone to changes, top management commitment, and the training of employees, and is built on three known pillars: those economic, environmental, and social.

An analysis of the barriers of green, lean practices in manufacturing industries [12] highlights the fact that company culture, top management commitment, and the training of employees are critical for both lean and green productions. Therefore, these three elements serve as the foundations of our model, as illustrated in Figure 1. On this foundation, we can start improving sustainability performance through the economic, environmental, and social dimensions of an organizational production system. The combined result of all three sustainability performances is green, lean sustainable production.

The process which a company should go through for lean, green sustainable improvement is shown in Figure 2. This process ends when appropriate measures have been taken and the process has improved, but this can also be a new starting point for subsequent new improvements, of course, if the process is determined to need improvement.

First, a company needs to identify the process to be analyzed. Starting from this decision, quality input data have to be collected, which will be used for green lean value stream mapping (GLVSM), which is the basis for all subsequent steps: the identification of opportunities for improvement and the elimination of waste and a reduction in environmental burden. The green value stream mapping (GVSM) approach, introduced by the authors in [34], expands the standard value stream mapping by incorporating additional flows such as energy, material, transport, and emissions, alongside the standard time stream. In our approach, we have further developed green lean value stream mapping (GLVSM) by integrating cost [35] and added value metrics, as illustrated in Figure 3.

Based on the GLVSM data, a radar chart diagram (Figure 4) can be drawn to see the current situation and to set goals for an improved process. Different solutions can be proposed which lead to different outcomes and different investments in improvements, considering the sustainable and environmental impact. Before implementing any changes, the return on investment (ROI) needs to be checked in comparison with reducing the environmental impact (EI). If it is accepted, the process should be improved and new input data gathered. If the ROI vs. EI is not acceptable, the company should find new solutions towards improvement.

In Figure 3, MRP represents production planning and control, which is part of Enterprise Resource Planning (ERP) in larger companies, but it is also an independent unit in smaller companies.

The innovative advantage of this model is the integrated cost and added value metrics, which represent the addition of all necessary information and values for the calculation of target functions FAP_j and FP_j, in addition to individual ratios in GLVSM.

Based on our model, we need to find the minimum of the first target function (FAP_j) of the accompanying processes such as storage, waiting, and transport, which do not add value, but some of which are necessary to perform the following processes:

$${\mathit{FAP}}_j=(\mathit{min}{\sum }_{k=1}^m{LT}_k+min{\sum }_{k=1}^m{EU}_k+min{\sum }_{k=1}^m{TMU}_k+min{\sum }_{k=1}^m{TTP}_k+min{\sum }_{k=1}^m{TCE}_k+min{\sum }_{k=1}^m{TSC}_k+min{\sum }_{k=1}^m{NVA}_k) ,$$

the jth version of the solution, and m, which is the total number of versions, and at the same time, the maximum of the second target function (FPj), which represents processes such as processing, assembly, and control, which add value:

$${\mathit{FP}}_j=(\mathit{max}{\sum }_{k=1}^m{TF}_k+max{\sum }_{k=1}^m{EF}_j+max{\sum }_{k=1}^m{MF}_k+max{\sum }_{k=1}^m{TRF}_k+max{\sum }_{k=1}^m{EMF}_k+max{\sum }_{k=1}^m{CF}_k+max{\sum }_{k=1}^m{VF}_k),$$

where:

-

Time flow ratio: ${TF}_j={\displaystyle \frac{{PT}_j}{{LT}_j}}$; ${PT}_j={\sum }_{k=1}^z{PT}_k$; ${LT}_j={\sum }_{i=1}^n{LT}_i+{\sum }_{k=1}^z{PT}_k$;

-

Energy flow ratio: ${EF}_j={\displaystyle \frac{{PEU}_j}{{EU}_j}}$; ${PEU}_j={\sum }_{k=1}^z{PEU}_k$; ${EU}_j={\sum }_{i=1}^n{EU}_i+{\sum }_{k=1}^z{PEU}_k$;

-

Material flow ratio: ${MF}_j={\displaystyle \frac{{MN}_j}{{TMU}_j}}$; ${MN}_j={\sum }_{k=1}^z{MN}_k$; ${TMU}_j={\sum }_{i=1}^n{TMU}_i+{\sum }_{k=1}^z{MN}_k$;

-

Transportation flow ratio: ${TRF}_j={\displaystyle \frac{{GP}_j}{{TTP}_j}};{GP}_j={\sum }_{k=1}^z{GP}_k$; ${TTP}_j={\sum }_{i=1}^n{TTP}_i+{\sum }_{k=1}^z{GP}_k$;

-

Emission flow ratio: ${EMF}_j={\displaystyle \frac{{VACE}_j}{{TCE}_j}}$; ${VACE}_j={\sum }_{k=1}^z{VACE}_k$;

-

${TCE}_j={\sum }_{i=1}^n{TCE}_i+{\sum }_{k=1}^z{VACE}_k$;

-

Cost flow ratio: ${CF}_j={\displaystyle \frac{{PC}_j}{{TSC}_j}}$; ${PC}_j={\sum }_{k=1}^z{PC}_k$; ${TSC}_j={\sum }_{i=1}^n{TSC}_i$;

-

Value flow ratio: ${VF}_j={\displaystyle \frac{{VA}_j}{{NVA}_j}}$; ${VA}_j={\sum }_{k=1}^z{VA}_k$; ${NVA}_j={\sum }_{i=1}^n{NVA}_i$

Explanations of the symbols in the equations:

TFj	Time flow ratio [-]
LTj	Production lead time [s]
PTj	Processing time [s]
EFj	Energy flow ratio [-]
EUj	Total energy used [W]
PEUj	Energy used for producing [W]
MFj	Material flow ratio [-]
TMUj	Total material used [kg]
MNj	Material needed [kg]
TRFj	Transportation flow ratio [-]
TTPj	Total transported pieces [m]
GPj	Good pieces transported [m]
EMFj	Emission flow ratio [-]
TCEj	Total carbon emission [1/kgCO2e]
VACEj	Value-added carbon emission [1/kgCO2e]
CFj	Cost flow ratio [-]
TSCj	Transport and storage cost [€]
PCj	Processing cost [€]
VFj	Value flow ratio [-]
NVAj	Non-value added [€]
VAj	Value added [€]

The target functions FAP_j and FP_j can be expressed in tropical notations by ${FAP}_j=\left(\oplus \mathrm{\text{'}}{\odot \mathrm{\text{'}}}_{k=1}^m{LT}_k\right)\odot \mathrm{\text{'}}\left(\oplus \mathrm{\text{'}}{\odot \mathrm{\text{'}}}_{k=1}^m{EU}_k\right)\odot \mathrm{\text{'}}\left(\oplus \mathrm{\text{'}}{\odot \mathrm{\text{'}}}_{k=1}^m{TMU}_k\right)\odot \mathrm{\text{'}}\left(\oplus \mathrm{\text{'}}{\odot \mathrm{\text{'}}}_{k=1}^m{TTP}_k\right)\odot \mathrm{\text{'}}\left(\oplus \mathrm{\text{'}}{\odot \mathrm{\text{'}}}_{k=1}^m{TCE}_k\right)\odot \mathrm{\text{'}}\left(\oplus \mathrm{\text{'}}{\odot \mathrm{\text{'}}}_{k=1}^m{TSC}_k\right)\odot \mathrm{\text{'}}\left(\oplus \mathrm{\text{'}}{\odot \mathrm{\text{'}}}_{k=1}^m{NVA}_k\right)$ and ${FP}_j=\left(\oplus {\odot }_{k=1}^m{TF}_k\right)\odot \left(\oplus {\odot }_{k=1}^m{EF}_k\right)\odot \left(\oplus {\odot }_{k=1}^m{MF}_k\right)\odot \left(\oplus {\odot }_{k=1}^m{TRF}_k\right)\odot \left(\oplus {\odot }_{k=1}^m{EMF}_k\right)\odot \left(\oplus {\odot }_{k=1}^m{CF}_k\right)\odot \left(\oplus {\odot }_{k=1}^m{VF}_k\right),$ respectively. Other values can be expressed in tropical terminology similarly. For example, the time flow equals ${TF}_j={{PT}_j}^{\odot {\displaystyle \frac{1}{{LT}_j}}}.$

Weights can also be added to the function (FP_j) to emphasize the importance of the company’s objectives.

Using the results of FAP_j and FP_j and the ROI_j, which are based on the investment and the company’s own price of the product, a decision about the optimum solution to be introduced can be made.

For visualization purposes, we suggest using the radar chart diagram, which illustrates the current situation and shows the direction of improvement and also the new situation after the improvement.

The radar chart shows the direction of our priority in the next optimization step.

The use of GLVSM and the radar chart diagram provides a better overview of the entire process, as well as a better orientation of the current situation of the company, and helps to determine the direction which the company wants to pursue in the future. After all, companies often take small steps towards their desired goals in terms of continuous improvement. A disadvantage of the model is that it is time-consuming and difficult to obtain all the data, because some companies have their databases organized much better than others.

4. Results

Based on the proposed methodology, the model was tested in a Slovenian company from the automotive sector.

First, the GLVSM was drawn (Figure 5) on the basis of the current situation (which was named Example 1). The production of this part consists of two technological operations, namely turning and grinding. The company’s output is 1870 pieces per week. To be on the safe side, a series included 3740 pieces. Based on these data, a recalculation was made for all items in the GLVSM diagram.

For example, for the time flow ratio:

$${TF}_1={\displaystyle \frac{{PT}_1}{{LT}_1}}={\displaystyle \frac{78.28}{\mathrm{1,900,014.28}}} =4.12\times {10}^{-5}$$

$${PT}_1={\sum }_{k=1}^2{PT}_k=40.42+37.86=78.28 \mathrm{s}$$

$${LT}_1=\sum _{i=1}^3{LT}_i+\sum _{k=1}^3{PT}_k=\left(\mathrm{633,312}+\mathrm{633,312}+\mathrm{633,312}\right)+\left(40.42+37.86\right)=\mathrm{1,900,014.3} \mathrm{s}$$

All other elements in the GLVSM were calculated using the same procedure (on the basis of the given equations in the proposed model section).

After creating the GLVSM for Example 1 and performing the calculation based on the equations in the proposed model section, the target functions FAP_j and FP_j can be calculated using the equations provided in the section on the theory.

FAP₁ = (min(633312 + 633312 + 633312) + min(56100 + 32300 + 58200) + min(4488 + 3665 + 3646) + min(25 + 10 + 30) + min(6.2 + 2.48 + 7.44) + min(193.5 + 102.3 + 207.5) + min(231 + 112.46 + 255.3)) = 2059596.5

FP₁ = (max(4.12 × 10 − 5) + max(0.0004911) + max(0.0001657) + max(0.0307692) + max(0.9888337) + max(0.0047685) + max(0.0133609)) = 1.03843

Based on this calculation, the minimum and maximum cannot be chosen, since only one value for each is available. It is only when several variants are available that a better variant or example can be chosen, based on the minimum value for FAP_j and the maximum value for FP_j.

This is the reason why some changes will be introduced in the process described in Example 1; this will be Example 2. The stock level will be changed from 7.33 days (633,312 s) to 4 days (345,600 s) and the energy used for transportation will be reduced by rearranging the layout of the machines in the production line and thus reduce the distance between the machines and the warehouses from the existing 25, 10, and 30 m (in Example 1) to 5 m between each machine and the warehouse in Example 2. This will also reduce the transportation costs, which also affect the other ratio calculations shown in the GLVSM for Example 2. For an easier and direct comparison of the ratio calculation between Example 1 and Example 2, the results are shown in a common table (

Table 1. Calculated data for radar chart diagram.

	TF [-]	LT [s]	PT [s]	EF [-]	EU [W]	PEU [W]	MF [-]	TMU [kg]
Example 1	4.12 × 10−5	1,900,014.3	78.28	0.000491	146,600	72	0.000165	11,799
Example 2	0.001812	43,200	78.28	0.004911	14,660	72	0.000166	11,799
	MN [kg]	TRF [-]	TTP [m]	GP [m]	EMF [-]	TCE [1/kgCO2e]	VACE [1/kgCO2e]	CF [-]
Example 1	1.955	0.030767	65	2	0.988834	16.12	15.94	0.004768
Example 2	1.955	0.133333	15	2	1.224270	13.02	15.94	0.007201
	TSC [€]	PC [€]	VF [-]	NVA [€]	VA [€]		FAP [-]	FP [-]
Example 1	503.3	2.4	0.0133609	598.76	8		2,059,596.5	1.03843
Example 2	333.3	2.4	0.0222990	358.76	8		70,379.1	1.39399

), which also includes the new values of FAP_j and FP_j.

As shown in Table 1, a change in production in one place can trigger changes (positive or negative) in the entire production chain in a company, which are clearly visible in this proposed model. Only the sum of all the changes in the form of FAP_j and FP_j values provides the final picture—a change in the production either brings gain or loss. The results of the individual variables are best displayed in a radar chart diagram (Figure 6):

If the diagram and the data are analyzed, it becomes apparent that in our case, time flow, energy flow, transportation flow, and cost flow are higher, while material flow, emission flow, and value-added ratio have remained the same. It became clear that Example 2 clearly reduced the environmental impact (EI) and that the ROI₂ calculated by the company was 25.3% (the ROI₁ was not calculated because Example 1 was the current situation).

If the values from Examples 1 and 2 are compared, Example 2 is found to be better because FAP₂ has a lower value than FAP₁, and FP₂ has a higher value than FP₁.

Based on all of these data and calculations, the company opted for Example 2.

5. Conclusions and Discussion

Based on an extensive literature review and our experience with integrating green and lean principles in a business context, we propose a new model and a step-by-step framework for sustainable green and lean business improvement along with performance measurement. This model provides a comprehensive roadmap outlining the necessary steps to achieve an environmentally focused and lean manufacturing practice. The proposed process facilitates a systematic pursuit of sustainable improvement that guides stakeholders through a progression sequence.

Using green lean value stream mapping (GLVSM), the user gains the ability to calculate two target functions (FAP_j and FP_j) and visualize the results in a radar chart diagram. A preliminary evaluation of these target functions, together with the consideration of return on investment (ROI) and environmental impact (EI) reduction, allows the user to make informed decisions about the implementation of the proposed changes.

By introducing the proposed model into the process, stakeholders can increase production levels while meeting measurable targets, moving towards a sustainable green lean production paradigm. In addition to the targeted improvements, the methodology emphasizes the assessment dimensions of the existing situation of production and the subsequent value after the improvements, providing a comprehensive understanding of the transformative impact on the production site.