Holistic Sustainable Design: Incorporating Change Propagation and Triple Bottom Line Sustainability

Abstract

Sustainable product design addresses sustainability challenges through product development processes and tools. The number of sustainable design methods has been increasing rapidly in recent years. Still, their adoption is limited, and many of these methods exclusively focus on the environmental impacts of products rather than taking a holistic perspective that includes social and economic sustainability. This research provides a holistic sustainable design framework by integrating change propagation methods and sustainable design tools to enable simultaneous consideration of design parameters’ impacts on the three dimensions of sustainability. A reusable water bottle is used to demonstrate the application of the proposed holistic sustainable design (HSD) framework. A multi-domain matrix (MDM) is used to capture the interdependencies among these design parameters of the product, and then equations are defined to quantify them. Life cycle assessment (LCA) is then automated to evaluate the product’s environmental impacts, and the investigation of its results provides details to identify critical unit processes contributing to the environmental categories. Sensitivity analyses reveal how changes to individual design parameters propagate through the model to influence the three dimensions of sustainability. Ultimately, the designer can select optimal design parameter values to balance environmental, social, and economic sustainability.

1. Introduction

Due to growing environmental concerns, achieving sustainability through sustainable design has gained attention in recent decades. Many industries are actively utilizing tools and methods to implement sustainability within their organizations, and the rise in societal awareness toward a sustainable future has pushed international agencies to adopt policies to protect the environment. The triple bottom line (TBL) concept defines sustainability with a holistic perspective, where it is sought considering people, planet, and profit [1]. Any system’s primary objective to achieve sustainability is to maximize positive social and economic factors while minimizing the negative environmental impact. Specific value forms included in each of these three dimensions of sustainability are shown in Figure 1. The main focus of sustainable design research is to provide practical product development support that can be utilized to achieve a sustainable future by systematically addressing these forms of sustainable value.

The United Nations (UN) 2016 report highlights 17 sustainable development goals (SDGs) that follow the TBL concept where sustainable product design can directly and indirectly contribute toward achieving these goals [2]. Design research holds great promise in driving advancements in products and systems, playing a pivotal role in pursuing the SDGs. Specifically, SDG 12, which addresses “sustainable consumption and production”, is relevant in product design. This goal calls for sustainable practices by both companies involved in product development and their consumers, necessitating policies and global agreements to ensure responsible management of environmentally hazardous materials. Researchers have pointed out that 80% of products’ sustainability impact could be considered and managed in the design phase [3,4]. The significant challenges in sustainable design emerge while balancing the three sustainability dimensions and product functionality needs [5]. Prior research highlighted the definition of sustainable product design, where it achieves its main function while ensuring the lowest environmental impacts and providing economic and social benefits to the stakeholders [6].

Figure 1. The triple bottom line (TBL) value forms [7].

Figure 1. The triple bottom line (TBL) value forms [7].

Designers have used design for X (DFX) methods to comprehensively consider product life phases and features (e.g., manufacturing, cost, assembly) during the design phase [8]. Design for environment (DFE) and design for sustainability (DFS) are the two categories of DFX that contribute toward incorporating sustainability in product design [9]. There has been a recent increase in the development of methods and tools with similar design principles, such as eco-design, sustainable design methods and tools (SDMTs), and environmental assessment tools [6]. Several methodologies have contributed to sustainable product design in recent years; however, their acceptance and adoption among designers remain limited [9]. Despite the advancements made through these methods, there is a lack of a practical holistic framework for sustainable product design [10,11]. DFX approaches focus on one phase of the product life cycle at a time and tend to focus on economic objectives rather than environmental or social impacts. More than 600 SDMTs are available, but their inability to attract widespread adoption is attributable to a lack of business context in their methods [12].

Environmental assessment tools, such as life cycle assessment (LCA), have gained widespread adoption across various industrial sectors and research fields, serving to evaluate the environmental impacts of products. The ISO-14040 [13] series has standardized the implementation of LCA, enabling the quantification of environmental impacts throughout the entire life cycle of products. Although numerous software tools and techniques are available for LCA, product designers often encounter challenges in effectively utilizing it due to a lack of interdependencies between the product structure and process requirements within a specific life cycle [14]. Consequently, the LCA results fail to provide meaningful insights into the environmental impacts associated with design decisions in the product system, and therefore, designers fail to perform sustainability trade-off analysis. In addition, the economic and social aspects of sustainability are not covered in traditional LCA. More recently, researchers have proposed a life cycle sustainability assessment (LCSA) that can potentially combine life cycle costing (LCC) and social life cycle assessment (SLCA) to provide a multidimensional perspective, but this has limitations, including an inability to consider temporality issues [15].

To address the challenges in current methods and tools for sustainable product design, this study proposes a holistic framework that integrates change propagation and DFS tools. This framework encompasses a holistic approach to sustainable design by integrating a multi-domain matrix (MDM) and LCA. Building upon the design structure matrix (DSM), the MDM is a network modeling tool to identify interdependencies among system elements. The MDM is employed within the framework to detect and quantify the dependencies among design parameters in components, products, and sustainability domains. Simultaneously, LCA is employed to evaluate the environmental impacts of the proposed product design. This article presents a reusable water bottle case study to demonstrate the framework’s application, including breaking down its components and design parameters, mapping out dependencies using MDM, mathematically defining them, and implementing LCA to assess the implications of design decisions. This comprehensive framework offers a practical tool enabling product designers to make sustainable design decisions by considering their choices’ environmental, social, and economic impacts. By providing a holistic perspective, the framework empowers designers to consider the broader implications of their decisions and implement sustainable practices effectively. This article is a revised and expanded version of a paper entitled “An Integrated Holistic Approach Toward Sustainable Product Design Using Life Cycle Assessment” which was presented at ASME International Design Engineering Technical Conferences and Computers and Information in Engineering Conference, Boston, MA, 20–23 August [16].

2. Background

A brief background and literature of tools and methods utilized in sustainable product designs, change propagation, and life cycle assessment is included in the following section.

2.1. Sustainable Product Design

“To design is either to formulate a plan for the satisfaction of a specified need or to solve a specific problem” [17]. The identification of the problem is the first step in product design, then synthesis, analysis, optimization and evaluation to achieve the final design. This iterative process has been adopted in different engineering design domains and most of these design decisions are influenced directly by requirements such as cost, quality, and functionality [6,18]. The growing global and societal environmental concerns led to the increasing number of sustainable design tools (sustainable life-cycle design (SLD), DFX, eco-design, etc.) that are aimed at incorporating sustainability considerations within the requirements or design processes. While these approaches are designed to improve sustainability, many focus on isolated factors rather than adopting a holistic perspective. For instance, eco-design mainly addresses environmental concerns, while Design for Environment (DFE) concentrates on specific life-cycle phases like manufacturing or usage [19,20].

One of the most commonly used tools in sustainable design is LCA. This method evaluates environmental impacts across a product’s entire life cycle, from material extraction to disposal/recycling. However, LCA requires detailed data on materials, manufacturing processes, and disposal methods, making early-stage implementation challenging. Additionally, simplified LCA tools face limitations due to limited data, lack of guidance, and minimal social sustainability integration [21]. Researchers highlighted the need for an LCSA as a more comprehensive approach that considers environmental, economic, and social factors [22].

Considering the advances in sustainable design tools, multiple challenges that limit their wide adoption remain. Some key issues include:

• Navigating a complex landscape of tools: The quantity of sustainable design methodologies available makes it difficult for practitioners to identify the most effective approach for a given product or industry.
• Lack of universal applicability: Many sustainability tools are designed to address specific industries or design challenges, making them unsuitable for broader implementation across diverse engineering fields.
• Overemphasis on environmental factors: Most available tools prioritize environmental impact assessment, often overlooking critical economic and social dimensions of sustainability.
• Weak links between design decisions and sustainability outcomes: Many existing tools fail to translate sustainability assessments into actionable design recommendations, limiting their practical application.

Researchers advocate for a more holistic and adaptable approach to sustainable design to address these issues. Integrating multiple frameworks and tools can provide a more balanced perspective on sustainability instead of relying on isolated methodologies. Design decisions can become more informed and aligned with long-term sustainability goals By bridging gaps between environmental, economic, and social factors [23,24]. New sustainable design tools require a specific focus on developing decision-support systems that facilitate the seamless integration of sustainability considerations across all phases of product development.

2.2. Change Propagation

As a first step toward solving a complex design problem, it is essential to identify the dependencies between design parameters. A decomposition and mapping process must be performed for each product component and its design parameters. This provides the basis to quantify and evaluate how these dependencies cascade through the product when a change is required. Design change is a common phenomenon in engineering design and can be initiated for various reasons, including improving the design, rectifying flaws, or adapting the design to meet changing requirements [25,26]. Scholars have identified design change as a source of innovation, and managing design change is essential as it can impact any information generated during the product development process and occur before or after releasing such information [27]. Practitioners and researchers have recognized the significance of managing design change mainly because it propagates, meaning that changes in one aspect of design initiate further changes to ensure that the design works cohesively, and such propagation of design change can be challenging to anticipate [28].

One of the most common tools to demonstrate and analyze change propagation is the design or dependency structure matrix [28]. DSM provides a matrix representation of interdependencies and relationships between different components, modules, or subsystems of a product or system [29]. In a DSM, the rows and columns represent the individual components or parameters, and the cells indicate their dependencies and relationships. These dependencies may be functional, physical, or informational. DSMs are useful for identifying potential design issues, reducing design complexity, and optimizing the overall system architecture. The matrix format of DSMs makes them highly compact and scalable compared to network diagrams, enabling them to handle large and complex systems [30]. Analysis techniques applied to DSM include partitioning/sequencing and clustering that aim to either reduce feedback marks or find mutually exclusive subsets [29]. Algorithms can prioritize components based on user-defined dependencies or manually inputted system logic, thereby rearranging the overall matrix order through partitioning or sequencing [30]. Clustering may lead to design efficiencies in manufacturing depending on the product under consideration. In some studies, researchers have investigated the possibility of generating DSMs automatically through the evolution of product models, where DSMs are generated automatically by extracting data from computer-aided design (CAD) models [31]. Researchers have applied DSM-based methods to utilize change management for product development [32].

Previous research has integrated an augmented DSM with a simplified LCA [33]. In this method, the multi-criteria decision-making method was implemented to compare the environmental performance of design alternatives, and the Pareto method was used to select the most promising concepts. Wang et al. applied a DSM technique to obtain optimized module efficiency in their model [34]. Their research aimed to achieve the goal of green design efficiency by applying quality function deployment (QFD). Shoval used a multiplayer DSM for constructing a holistic view of the system’s architecture [35]. Their study uses a clustering algorithm to identify the possible modularization in each life cycle phase. In summary, various application and integration opportunities are available for using DSM in sustainable product design.

Cross-domain DSMs are the expansion of single-domain DSMs, allowing for the evaluation of design changes across multiple domains. These matrices are the design dependency matrix (DDM), domain mapping matrix (DMM), and MDM. MDM is known as “a square matrix comparable to a DSM containing system elements in identical order on both axes. In contrast to a DSM, different system elements are included and grouped in domains” [36]. The MDM includes all the features of the DSM but represents them on a higher level of abstraction and with greater flexibility to include elements representing different types of parameters or properties [37]. Koh et al. implemented an MDM to represent product components, options, and requirements as inputs for change propagation analysis [25]. This also applies to sustainable product design, mainly due to the expansion of design parameter dependencies within multiple domains. MDMs are found to be helpful for the simultaneous integration of different domains, but there are no available rules to reliability analyze them [36,38]. Bartolomei et al. proposed using an engineering system’s multiple-domain matrix (ES-MDM) as an organizing framework for modeling engineering systems that map interactions within and across domains and parameterize their interdependencies [39]. The problem of change propagation in the context of this study expands to different domains, and the dependencies between these parameters can be mapped using MDM across these domains.

Change propagation is an engineering design concept that evaluates how one component’s modifications can trigger cascading effects across interconnected subsystems. Traditionally, change propagation has been used in systems engineering and complex product development to manage design iterations and minimize unintended consequences [25,40]. However, its application in sustainable product design is still emerging. Integration of change propagation and sustainability assessment provides anticipated design decisions that impact environmental, economic, and social sustainability metrics. The designer can ensure that changes to optimize one sustainability dimension do not lead to unintended negative consequences for others utilizing this integrated approach.

2.3. Life Cycle Assessment

LCA is a widely used method that can model and calculate the environmental impacts of products and processes throughout their entire life cycle [41]. Applications of applying LCA in many industries and research have shown it to have various capabilities and limitations [15,22], and it can be used to inform decision-makers about the environmental impacts of their products. The International Organization for Standardization (ISO) has published ISO 14044 [13], which provides a detailed guideline about the standard practice of LCA. The overall process of LCA includes four steps: goal and scope, inventory analysis, impact assessment, and interpretation.

• Goal and Scope: The first step of LCA is to define the goal and scope of the analysis, which includes identifying the system boundaries and specifying the life cycle stages that should be included in the analysis. The product and product system should be described, including the product’s upstream and downstream processes. The product’s components, manufacturing, distribution, use, and disposal are considered in these streams. In addition, the product’s intended function and the functional unit should be defined in this stage. A functional unit is a measure used to quantify a product’s function and provide a standard basis for comparisons. Another parameter defined in this phase of LCA is the reference flow, which includes the product required to fulfill the defined function. Using the parameters in the previous steps, the system boundaries can be defined by selecting which unit processes need to be included in the analysis. The selection process is done based on the environmental significance of the processes. Then, the selected unit processes and boundaries are combined to create a process tree to demonstrate the inputs and outputs of the unit processes within the production system.
• Inventory Analysis: The life cycle inventory (LCI) analysis includes data collection, validation, and calculation of the inputs and outputs of the product system defined. These data include inputs (e.g., energy, materials, water) and outputs (e.g., emissions, waste) associated with each life cycle phase of the product. The sources of LCI are categorized into primary and secondary data. The primary data are collected by direct measurements and observations with the highest accuracy, but they are time-consuming and costly. Secondary data refers to data collected from existing sources that provide access to a wide range of information collected by various entities, making it easier to gather data efficiently. The activities and processes within the system boundary defined earlier are analyzed concerning a reference flow (functional unit) to determine the overall inputs and outputs of the product system.
• Impact Assessment: The inventory data collected in the LCI analysis step is implemented in the life cycle impact assessment (LCIA) to quantify the product’s environmental impacts. LCIA includes classification, characterization, and two optional steps of normalization and weighting. The inventory data collected are classified into different impact categories. The LCIA method selected for LCA is an essential part of the framework to represent the environmental impacts in a meaningful format. One of the most popular LCIA methods is ReCiPe, which transforms the LCI results into 18 midpoint categories and then aggregates those midpoints into three endpoint categories. These endpoints are also known as areas of protection, which are damage to human health, damage to ecosystems, and damage to resource availability [42]. Each endpoint is measured in terms of its specific unit representing the value of environmental impacts. The damage to human health is measured as disability-adjusted life years (DALYs), the unit for damage to ecosystems is the local species loss integrated over time (species.year), and the damage to resource availability has a unit of U.S. dollars (USD), which represents the extra costs involved for future mineral and fossil resource extraction. ReCiPe offers three cultural perspective factors: I, H, and E. The individualist (I) perspective is a short-term, optimistic view that technology can avoid many problems in the future. The hierarchist (H) is based on scientific consensus concerning the time frame, and the egalitarian (E) perspective takes a long-term view based on precautionary thinking.
• Interpretation: The last step of LCA is interpreting the results, which includes three key elements: identification of key issues, evaluation of the results, conclusions, and recommendations.

Despite LCA’s capabilities, designers face challenges in effectively using LCA tools as intended due to the absence of interconnections between product architectures and process requirements throughout a specific life cycle. The gap between LCA results and design decisions and the interdependencies between these decisions makes it challenging to implement LCA results in the design phase of product development. It has also been noted that the results obtained from LCA have high uncertainties concerning the required input data [14,41]. Environmental policies significantly influence engineering design decisions, and companies are trying to meet compliance requirements with the lowest cost [43]. LCA is limited in its utility in the design process, making it an inadequate tool for the designer. Researchers concluded that LCA could be considered a specialized tool handled by a special player (the environmental actor) and should be dedicated to strategically evaluating new concepts [44]. Ahmad et al. reviewed tools, applications, and research prospects published from 2007 through 2017 in their research [6]; they found that most sustainable product design (SPD) tools considered only limited environmental analysis rather than covering all three pillars of sustainability (environmental, social, and economic). Simplified LCAs have been integrated into some computer-aided design (CAD) software tools, but comparing their results to dedicated LCA revealed a high inaccuracy in results [45]. Researchers also propose a hybrid LCA method to combine the accuracy of process analysis and the completeness of input-output analysis [46]. The accuracy of hybrid LCA compared to process-based LCA has been debated among experts in this field [47,48].

For the non-environmental pillars of sustainability, separate methods have been defined and studied. Life cycle costing (LCC) is a method that has been applied to account for the economic impacts of a product throughout its life cycle [49]. Researchers have investigated the possibility of shifting from LCC to economic life cycle assessment (EcLCA), considering the interlinkage between the micro and macroeconomic levels [50]. In their method, areas of protection are economic stability and wealth generation, driven by two endpoint categories: economic prosperity and economic resilience. In addition, social life cycle assessment (SLCA) aims to assess the social impacts of products and services across their life cycle [51]. Researchers have proposed two options to capture the three-pillar approach to sustainability [52]. The first option is LCSA = LCA + LCC + SLCA, based on three separate assessments with consistent and identical system boundaries. The second option is LCSA = LCA new (including LCC and SLCA as additional impact categories in LCIA), which requires LCI to follow three impact assessments. LCSA methods aim to encompass all three pillars of sustainability following ISO 14040 and 14044 for environmental life cycle assessment, LCC for economic sustainability assessment, and SLCA for social sustainability [53]. LCSA can help provide a holistic assessment of sustainability for stakeholders and assist them in decision-making [54], but challenges arise when pairing two distinct tools (e.g., LCA, LCC, and/or SLCA) with one another. The significant challenges are temporal issues, different perspectives, and indirect consequences [15]. Methodologies for integrating LCA and LCC, which include multidisciplinary design optimization (MDO), have also been proposed [55].

Moreover, researchers used an artificial neural network (ANN) based LCA model to address the difficulties of performing full LCA [56]. They implemented hierarchical clustering and a decision tree algorithm for systematically identifying product groups according to their environmental categories. Other research has attempted to address these concerns by reviewing the challenges of simplifying LCA tools to produce higher levels of sustainability [21]; however, these do not explicitly solve the problem of integrating LCA into the earlier design phases. Poole et al. introduced an LCA matrix to summarize the product’s environmental impacts using a simplified LCA method [57]. De Napoli et al., in their methodology, used a simplified LCA and integrated it with an augmented DSM [33]. In this method, the multi-criteria decision-making method was implemented to compare the environmental performance of design alternatives, and the Pareto method was used to select the most promising concepts.

Altogether, LCA results are often too complex for designers not familiar and experienced with environmental assessments [58]. Since improving LCA databases requires significantly more data, regionalization information, and investment from all contributors, improving LCA techniques is a trade-off situation where improving the system could make it even more challenging to implement and use. These challenges with LCA tend to create hurdles among individual stakeholders while leading product designers to move towards more precise alternative tools to estimate environmental impacts on particular components or materials for a specific part [59].

3. Methodology

This article approaches the complex multi-dimensional problem of sustainable design by proposing a novel holistic sustainable design (HSD) framework. A comprehensive and quantitative understanding of environmental, social, and economic impacts caused by design changes of a product can be achieved by utilizing this framework. The goal is to enable iteration within this framework to achieve an optimum design parameter where the three dimensions of sustainability are balanced. This section includes a detailed description of the framework, case study, and the methodology’s critical phases, including system definition and change propagation.

An overall structure of the proposed HSD framework is shown in Figure 2, which is divided into four major phases: system definition, change propagation, sustainability assessment, and sustainability improvement. First, in the system definition phase, the system of interest and its requirements are specified and analyzed to generate an MDM containing manually identified design decision dependencies. The design parameter dependencies are used to evaluate the product attributes in the change propagation phase. These product attributes are inputted for LCA calculations to determine the product’s environmental impact. In the third phase of sustainability assessment, the environmental, social, and economic sustainability metrics are measured and analyzed to see if they meet the designer’s needs. If unsatisfied, the sustainability improvement phase is initiated, where the critical processes impacting the environment can be identified by further investigating the LCA results, and the MDM can provide a detailed understanding of design parameters associated with that process and dependencies of that parameter across different system domains. Sensitivity analysis can also reveal improvement areas within the design parameters, and optimization methods can be implemented to find the best values for each design parameter to improve sustainability and performance objectives. The results of this stage are used to revise the system definition elements and determine if system requirements or boundaries should change to achieve sustainability. The framework has been implemented using Python programming to streamline this process flow. The loop structure illustrated in Figure 2 highlights the iterative process required to achieve optimal value for the final product.

3.1. Reusable Water Bottle Case Study

A reusable water bottle was selected as a suitable physical product for examining the early application of this framework, shown in Figure 3. The product is relatively simple, enabling the enumeration of design parameters and dependencies, and its technical, environmental, social, and economic features are relevant to the goals of the HSD analysis. It is also a consumer-based product that ensures the existence of sufficient socio-economic factors. This product has three components: bottle, cap, and seal made of stainless steel 304, polypropylene plastic resin, and silicone rubber, respectively. The dimensional and geometrical details related to the design of the reusable water bottle were extracted from available online sources [60].

3.2. System Definition

The system definition consists of two types of data: system requirements and system boundaries. The system requirements include all the constraints defined for the system to satisfy the product’s objective, such as constraints for weight, volume, and cost of the reusable water bottle. The reusable water bottle features were defined by estimating maximum values for the system constraints. The system boundary is then generated considering the constraints identified in the required data. These data were mainly generated based on estimation and assumptions, which is sufficient for this analysis, as it is intended to serve as a proof-of-concept example rather than provide actionable water bottle design recommendations. However, the system definition phase is a critical step in this framework, as it lays the foundation for the change propagation model.

The system requirement and boundary data provide the principles for generating the MDM, which consists of three domains: components, product attributes, and sustainability. The CAD model and two-dimensional drawings identify the design decisions associated with each component (bottle, cap, and seal) in the reusable water bottle. Each component has its individual dimensional, material, weight, manufacturing, and cost parameters assigned to them, and in the product attributes domain are specifications where all the individual component level parameters combine to define specific product features such as volume and weight. The next level of parameters identified in the product attributes domain is related to environmental impact assessment, such as energy, transportation, and waste. The product attributes domain also consists of performance parameters associated with the reusable water bottle: heat efficiency, durability, portability, washability, cost, quantity, and usage. The final domain included in this MDM is the sustainability domain, which includes the environmental, social, and economic sustainability parameters. Overall, 43 parameters in this case study are assembled into an MDM to map the dependencies between each parameter.

A completed MDM generated using the design parameters of the reusable water bottle case study is demonstrated in Figure 4a. The dependencies were mapped manually using engineering data, including two-dimensional drawings, CAD models, and bills of material (BOMs) of the reusable water bottle. The MDM’s columns and rows contain the design parameters from component, product attributes, and sustainability domains, and the dependencies between them are identified using “1” marks. Following the rows in the matrix, the marks in each row represent the inputs required for that design parameter, and the columns show the output of each design parameter. These dependencies mapped across the MDM result in a dependency report that lists all the design parameters in each domain and their required inputs. This report is then utilized to quantify the dependencies and determine each parameter based on its dependencies. For readability, a sample section of the MDM is shown in Figure 4b, illustrating how bottle diameter, mouth diameter, thickness, and material are inputs for determining the weight of the bottle. Change propagation is enabled when these relationships are quantified for each dependent parameter pair within the MDM.

3.3. Change Propagation

The dependencies mapped in the MDM are used as the basis for identifying and quantifying changes and their impact in these three domains. The initial changes always occur in the component domain, propagating to the product attributes and ultimately to the sustainability domain. The design parameters were collected by exploring the product’s assembly, the interaction between components, and product attributes, such as performance parameters. Each component includes dimensions (e.g., height, diameter, thickness), materials (e.g., stainless steel, polypropylene resin plastic), cost (e.g., material, manufacturing), and weight. The parameters listed in the product domain are where the components go through a common final production phase, such as final assembly, transportation, and packaging. The system’s energy, recycling, and waste parameters are used for LCA inputs to complete the environmental impact analysis of the product, which are grouped by product attributes domain. The performance parameters are also included in the product domain, which is essential to determine the sustainability metrics in the assessment phase of the framework. The weight and volume are the critical parameters implemented to determine the cost, final price, expected quantity, and LCA inputs.

These dependencies are quantified by considering the existing relation between the design parameters. For example, the ratio between the bottle mouth diameter and bottle diameter is the basis of evaluating this dependency. For each unit of change in bottle diameter, the bottle mouth diameter, the seal diameter, and the cap diameter will be changed accordingly. It is important to note that there is also a diameter-to-thickness ratio for the bottle design, which means that the thickness needs to be increased after a certain diameter. Here, for each unit of change in diameter, thickness, or height of the bottle, the volume and weight would be calculated again (updating the component domain) and will automatically update the product attributes and sustainability domains. The dependencies were quantified to enable change propagation within the system boundary. The performance parameters in the product domain were determined based on input parameters (rows) identified in the MDM, and the process of calculating the mathematical relationships between parameters and each dimension of sustainability is explained in the following subsections.

3.3.1. Economic Sustainability

Economic sustainability is evaluated through a profit-based metric incorporating production costs, markup margins, and expected product demand. The total cost includes raw materials, manufacturing, packaging, and transportation, while the selling price is determined based on a fixed markup factor. The demand estimation considers product durability and quality, linking consumer preference to financial viability. This profit-driven approach ensures that sustainability improvements do not compromise economic feasibility. Validation was conducted by benchmarking cost and demand assumptions against market data for similar products. The economic aspect of sustainability is mainly influenced by the production cost of a reusable water bottle, C, calculated using Equation (1).

$$C=C_B+C_C+C_S+(MF·a)+(PK·b)+(TR·c)$$

(1)

where $C_B$, $C_C$, and $C_S$ are the bottle, cap, and seal costs. $MF$ and $PK$ represent the manufacturing and packaging of the product measured as volume, and a and b are constant costs associated with one unit of volume. $TR$ is the transportation of the product measured in metric tons-km, and c is a constant cost associated with transporting one unit of weight. The total cost of the reusable water bottle is used to calculate the product’s final price (P), from Equation (2), which is simplified by assuming a fixed 80-percent markup ($\alpha$ = 1.8). This accounts for marginal profits earned by the producer and markups taken by others, such as distributors and retailers.

$$P=\alpha ·C$$

(2)

The product’s durability ($DU$) is measured as the number of years the product lasts, and assuming the material is stainless steel 304 and the bottle thickness ($T_B$) is less than 100 mm, durability is calculated by Equation (3).

$$DU=10·\sqrt{T_B}$$

(3)

The level of consumer demand in the market determines the quantity of water bottles expected to be sold. The Equation (4) defines how quantity sold (Q) increases with product quality and decreases with the price (P) in this example problem.

$$Q=(0.5·{\displaystyle \frac{DU+HE+PO}{3}})+(0.5·{\displaystyle \frac{1000}{P}})$$

(4)

The economic aspect of sustainability is represented by profit $\pi$, calculated by Equation (5).

$$\pi =Q·(P-C)$$

(5)

3.3.2. Social Sustainability

The social sustainability metric is derived from three key factors: job creation, product usage, and human health impacts. Job creation is estimated based on the number of manufacturing and transportation processes, reflecting the labor demand generated by production. Product usage is measured through heating efficiency, portability, and washability, representing consumer interaction and long-term adoption. LCA quantifies human health impacts, a factor that quantifies the total negative impact on the population’s health. The quality (q) parameter of the product is a critical factor in determining the usage of the product, and it is defined in Equation (6) as an average of three factors.

$$q={\displaystyle \frac{HE+WA+PO}{3}}$$

(6)

Here, $HE$ is heating efficiency, which is a function of the thickness and material of the bottle, and $WA$ is the product’s washability, which is a function of the bottle’s mouth diameter. The $PO$ is the portability of the reusable water bottle, which is a function of the bottle diameter and the product’s total weight. All three attributes are defined within a normalized range of 0 to 100, where the perfect water bottle would have a quality of 100.

The reusable water bottle usage (U) is determined as the number of times the customer uses the product in a year, where Equation (7) prescribes a linear relationship between quality (q) and usage. This formula enables the definition of the water bottle used by the customer in the product’s use phase into LCA analysis, as well as the number of functional units per product. The assumption is that the highest-performing water bottle would be used an average of once daily.

$$U=\left({\displaystyle \frac{q}{100}}\right)·365$$

(7)

Another social aspect associated with the product is the number of jobs created (J) for its production, calculated by Equation (8).

$$J=(N_C+N_M+TR)·{\displaystyle \frac{Q}{1000}}$$

(8)

Here, $N_C$ is the total number of components in the product, $N_M$ is the total number of manufacturing processes, and $TR$ is the mass of product transported multiplied by distance. In the context of this analysis, these factors can provide an estimated value of jobs created to evaluate the social aspects of sustainability based on product design.

Finally, social sustainability is represented by defining a composite social metric of the product (S), scaled from 0 to 100. Equation (9) is used to calculate the social metric for social sustainability of this reusable water bottle, where number of uses in a year (U), number of jobs created (J), and damage to human health (H) are weighted and divided by their maximum estimated value. These maximum estimated values are obtained based on assumptions relevant to this example and can be changed in different scenarios. The damage to human health is calculated as disability-adjusted life years (DALYs), obtained from the LCA calculation described in the next section. Each of the three social metrics is weighted with subjective values of 0.3 or 0.4, which can be set or adjusted by the design team.

$$S=(0.3·{\displaystyle \frac{U}{365}}+0.3·{\displaystyle \frac{J}{500}}-0.4·{\displaystyle \frac{H}{1}})·100$$

(9)

3.3.3. Environmental Sustainability

The environmental sustainability is calculated using LCA, following the four steps described in Section 2.3. The following sections describe the details of the goal and scope, life cycle inventory, life cycle impact assessment, and interpretation of the LCA performed for the reusable water bottle.

Goal and Scope

This analysis aims to reveal the environmental impacts of the reusable water bottle. The LCA results will be used as a baseline to compare against potential design changes. The product system is a reusable water bottle made of a stainless steel 304 bottle, polypropylene plastic resin cap, and silicone rubber seal. The scope is a cradle-to-grave analysis that contains the product’s material extraction, components manufacturing, transportation, usage, and end-of-life. The function of this product is to be used as a container for liquids, and the functional unit selected for the reusable water bottle is 1000 L of water stored and transported. Twelve unit processes are defined within the boundary of the system selected for this product, including three materials (stainless steel 304, polypropylene plastic resin, and silicone rubber), three manufacturing processes (steel production, powder coating, injection molding), electricity, packaging, transportation, water used, stainless steel, and plastic disposal to landfill. The total lifespan of the product (L), representing the total number of uses before the end of life, is calculated by Equation (10).

$$L=U·DU$$

(10)

The functional unit (F) of LCA analysis (1000 L of water stored and transported) per product is calculated by Equation (11), where V is the internal volume of liquid the product contains.

$$F={\displaystyle \frac{V·L}{1000}}$$

(11)

Life Cycle Inventory

The technology matrix (A) is generated using the input and outputs of the product system generated based on unit processes from the goal and scope, and the values assigned to each flow are determined using the equations defined based on the MDM’s dependency report. The functional unit of the product is used to generate a reference flow matrix (demand matrix) f. Since A and f are known, the scaling factor matrix (s) is evaluated using Equation (12) [61].

$$s=A^{-1}·f$$

(12)

The life cycle inventory (LCI) data are then represented as the environmental intervention matrix, B. These data are collected using the Ecoinvent database [62], and then the inventory matrix (g) is calculated using Equation (13).

$$g=B·s$$

(13)

Life Cycle Impact Assessment

The characterization factors (CF) matrix ($Q_c$) is obtained from the ReCiPe 2016 midpoint hierarchist (H) LCIA method [42]. After the inventory matrix (g) is classified into each category for each unit process, the environmental impact matrix (h) can be calculated using Equation (14).

$$h=Q_c·g$$

(14)

The matrix h consists of 18 midpoint indicators further classified into the three endpoint categories from ReCiPe. The CFs for three endpoint categories were also implemented using the hierarchist perspective [42]. The endpoint categories aggregate all the categories into three main areas of protection: DALY, species.year, and USD. Then, by evaluating the maximum midpoint category in each protection area, it is possible to determine the unit process with the highest contribution to each category.

The normalization and weighting are optional steps that can be used to further aggregate these impacts into single score points, measured as the average world citizen’s annual share of environmental impacts in 2010, using single score eco-indicator points (Pt) [63]. The weighting method in this LCA also follows a hierarchist perspective, where the ecosystem endpoint category weights 400, whereas human health and resource weights are 300. Considering the limitations and uncertainties associated with applying normalization and weighting in LCA analysis, it is helpful to determine a single score value for environmental sustainability to be represented in the sustainability assessment phase alongside the social and economic dimensions. The interpretation of the LCA results is demonstrated in the results and discussion.

4. Results

The results obtained from implementing the HSD framework for the reusable water bottle case study are presented in this section. The first round of results was obtained by inputting the default values assigned to the product’s design parameters to establish a baseline for the analysis. This section describes the sustainability assessment and improvement phases of the HSD framework.

4.1. Sustainability Assessment

The set of equations defined in Section 3.3, using the dependencies from the MDM, resulted in three quantifiable dimensions of sustainability.

Table 1. Sustainability assessment of baseline reusable water bottle.

Sustainability Dimension	Metric	Value
Environmental	Single Score (mPt)	9.06
Social	Social Metric (%)	38
Economic	Profit (USD)	466,881

shows the environmental, social, and economic sustainability indicators and the values obtained by implementing the proposed HSD framework on the baseline design. One single score point (Pt) represents the average world citizen’s annual share of the environmental impacts in 2010. The social and profit indicators are obtained by Equations (5) and (9), respectively. To comprehensively analyze the baseline results, further investigation into the environmental sustainability metric is required to understand which life cycle phases and processes contribute most to the environmental impacts.

The results of the LCA can be analyzed by taking a step backward from the single score to endpoint indicators and from there to midpoint categories.

Table 2. Most significant midpoint category for each endpoint category, from ReCiPe LCA results of baseline reusable water bottle.

Endpoint	Total Value	Most Significant Midpoint
Damage to Resource availability	$3.03\times {10}^{-3}$ USD2013	Mineral resource scarcity
Damage to Human health	$7.49\times {10}^{-6}$ DALY	Global warming
Damage to Ecosystems	$1.51\times {10}^{-8}$ species.yr	Global warming

shows the most significant midpoint category for each endpoint category, extracted from the ReCipe LCA results of the reusable water bottle baseline. The damage to resource availability shows that mineral resource scarcity has the highest contribution to this endpoint, and global warming has the highest contribution to damage to human health and the ecosystem. The most significant unit process contributing to these midpoint categories can be traced using the LCA midpoint results. Stainless steel, the bottle material, has the highest environmental impact on mineral resource scarcity and global warming. Considering the product’s main component is the bottle, which has the highest weight, it is logical that this process has the highest impact.

The midpoint results can be further explored to understand the environmental impact associated with each life cycle phase of the product. This is shown in Figure 5, where the environmental damage of each life cycle phase of the reusable water bottle and their unit processes are illustrated. According to the results, the material extraction phase accounts for 10% of the total environmental impacts, and stainless steel has the highest percentage of environmental damage compared to the other two unit processes. Focusing on the manufacturing phase (38%), powder coating and steel machining have the highest environmental damage after electricity, which is used to produce the product. The impact of the shipping is minimal (2%). Still, considering the functional unit of the LCA analysis (1000 L of water stored and transferred over the product’s lifespan), water used during the product’s use phase has high environmental damage (16%). The life cycle phase with the highest environmental damage is the end of life, with a 65% contribution to the product’s total environmental damage, and the unit processes associated with the stainless steel bottle material have the highest environmental damage compared to other processes. This LCA analysis assumes plastic and steel components are landfilled (cradle-to-grave).

These findings also underscore the importance of recycling or reusing stainless steel bottles, considering their significant environmental impact in the end-of-life phase. The designer can also explore available mechanisms to turn this analysis from cradle-to-grave to cradle-to-cradle, following the trend toward a circular economy.

4.2. Sustainability Improvement

The sustainability improvement is executed based on the assumption that the results obtained from the sustainability assessment may not be ideal for the product. The improvement phase can be initiated by identifying the area of improvement within the product design parameters. The sustainability assessment results showed that the end-of-life phase of the reusable water bottle has the highest contribution. Within that phase, the stainless steel bottle material is the critical process contributing to the total environmental impact percentage. Critical design parameters, such as durability, heating efficiency, and portability, are the criteria for selecting stainless steel as the bottle material. Two possible improvements can be explored about bottle material: (1) changing the bottle material and (2) changing the bottle weight. The baseline HSD analysis summarized in Table 1 has been performed using stainless steel 304 as the bottle material, and one approach to improving the environmental impact would be to change that to polyethylene terephthalate (PET) plastic, a well-known alternative to steel bottles. The second approach is to reduce the amount of stainless steel, calculated based on the bottle dimension parameters. According to the MDM dependency report, input design parameters associated with bottle weight are bottle height, neck height, bottle diameter, mouth diameter, thickness, and material. These input parameters are potential sources of change to reduce the bottle weight and the product’s environmental impact. The following sections include investigating these two possible improvements and their results.

The change in bottle material propagates to important product attributes such as weight and durability. Figure 6 compares the bottle materials and their sustainability impacts. It is essential to note that the LCA performed for the PET plastic was also cradle-to-grave, but the change of bottle material made essential changes within the system’s boundary. The manufacturing process associated with the bottle was changed to injection molding, the powder coating process was eliminated, and the landfill case was changed to only plastics. Moreover, the total amount of electricity used to manufacture the product is also changed to provide a fair comparison between the two scenarios.

The results obtained from this analysis demonstrate that the environmental impact of the reusable water bottle would increase by changing the material from stainless steel 304 to PET plastic by 1.69 mPt. PET plastic has lower weight and does not require a powder coating process, but it has lower durability and requires more electricity. Therefore, considering the functional unit of the analysis, stainless steel 304 is the desirable material, considering the environmental impact. Moreover, the holistic perspective provided within this framework shows that economic and social sustainability benefits will be reduced by using PET plastic as the bottle material. 85,600 USD reduces the profit earned, and the social sustainability metric is reduced from 38% to 31%. This is mainly due to the decreased usage of reusable water bottles with lower-durability plastic material. Overall, stainless steel 304 remains the better choice regarding the bottle material, considering environmental, social, and economic sustainability. This step within the HSD framework provides the designer with the foundation for informed decision-making and scenario analysis.

Next, the second improvement area is investigated by performing sensitivity analysis on three main design decisions associated with the bottle weight: bottle diameter, bottle mouth diameter, and bottle height. The lower and upper boundaries of each variable are defined within the system definition phase, and each sensitivity analysis aims to provide a quantitative exploration of the impact of design decision change on the three dimensions of sustainability.

The baseline value for the bottle diameter is 63.5 mm, and Figure 7 demonstrates how decreasing or increasing the bottle diameter impacts metrics for social (S), economic ($\pi$), and environmental (mPt, from ReCiPe single score) sustainability. Figure 7 contains three separate vertical axes, as each line has a different unit of measurement specified with colors. The objectives are to maximize the social and economic impacts while minimizing the environmental impacts. The changes in bottle diameter propagate through the component domain first, and the primary product attributes changing with bottle diameter are bottle weight and volume, which directly impact sustainability metrics. As shown in Figure 7, a decrease in bottle diameter (below 55 mm) causes a rapid decrease in sustainability’s economic and social dimensions. In addition, environmental sustainability increases by reducing the bottle diameter due to the low volume of the bottle and, therefore, fewer functional units provided per fill. Thus, lowering the bottle diameter below 50 mm is not desirable for sustainability metrics.

The profit peaks at 70 mm diameter at 467,000 USD, rapidly decreasing to 451,000 USD at 85 mm. Social sustainability follows a similar pattern: peaking at 70 mm with 39.3% and dropping to 38.6% at 85 mm. Moreover, the environmental damage decreases by increasing the bottle diameter within the allowable range. The baseline result for the environmental metric was 9.06 mPt at 63 mm, and it is lowered to 8.65 mPt at 85 mm due to its capacity to store more water over its lifespan. The findings show an acceptable range between 55 to 70 mm for bottle diameter, considering social and economic sustainability metrics. Moreover, any larger bottle diameter is more desirable to minimize the product’s environmental impact.

Figure 8 demonstrates observations across different bottle mouth diameters, enabling comprehensive identification of change propagation trends in each sustainability dimension. The range of the bottle mouth diameter spans from 30 to 70 mm. According to the results, a gradual increase is detected in economic sustainability as the bottle mouth diameter increases from an initial value of 443,000 USD at 30 mm diameter to 467,000 USD at 48 mm. Then, the profit decreases to 446,000 USD at 70 mm, mainly due to the low utility of the product because of its high volume and weight.

A slight upward trend can be observed in the social sustainability percentage as the bottle mouth diameter increases from 30 mm, and it peaks at 55 mm with 39.2%. Then, it gradually decreases to 38.5% social impact at 70 mm diameter due to the product’s low utility with this design. The environmental sustainability results obtained from LCA reveal the environmental impact associated with different bottle mouth diameters. It displays a downward trend as the diameter increases. Beginning at 10.69 mPt for a 30 diameter, the value decreases to 8.54 mPt for a 70 diameter due to the bottle’s ability to carry a higher volume of water over its lifespan. Overall, a specific range between 48 mm and 55 mm can maximize both social and economic sustainability metrics, and any larger bottle mouth diameter will continue to lower the environmental damages within the range of this analysis.

The bottle height is the third design decision in the product domain that can effectively impact the bottle weight and reduce the product’s environmental impact. Figure 9 explores the relationship between bottle height and its sustainability effects. The bottle height ranges from 100 to 300 mm, and the results show positive social and economic sustainability trends by increasing the bottle height.

Starting from 436,000 USD at a height of 100 mm, the profit gradually rises to 474,000 USD for a bottle height of 300 mm. This trend implies that taller bottles are associated with higher economic sustainability. As the bottle height increases, social sustainability exhibits an overall upward trend. Starting at 36.2% at a height of 100 mm, it reaches its peak of 39.2% at 300 mm, showing a similar trend to economic sustainability. However, the environmental damage increases incrementally by increasing the bottle height. Notably, reducing the bottle height below 150 mm also increases the environmental impacts due to lower volume and total functional units provided over the bottle’s lifespan. Still, the variation of these changes is minimal compared to previous results. Overall, increasing the bottle height to its upper bound would have positive social and economic sustainability impacts while increasing the environmental impacts from 9.06 mPt (baseline) to 9.25 mPt.

5. Discussion

The HSD framework put forward in this article is a new approach to late-stage design that enables designers to evaluate and improve the sustainability of their products. Through a relatively simple case study, the results demonstrate how the HSD framework can be used to find significant sources of environmental impacts and explore the impacts of different design parameters. This section highlights a more extensive explanation of the findings by discussing the results of the sustainability improvement phase and the overall application of the HSD framework.

5.1. Case Study Results

The case study results identified the two main parameters that could effectively reduce the environmental impact of the reusable water bottle: bottle material and bottle weight. While these findings may be self-evident or intuitive for a simple product, this analysis can reveal meaningful insights into a more complex product system where multiple components and materials are available. The sustainability improvement phase of the HSD framework was demonstrated by investigating changing the bottle material and performing a sensitivity analysis of the design parameters. The MDM then revealed the interdependencies associated with the bottle material and weight. Moreover, the design space has been explored through sensitivity analysis of the three design parameters associated with the bottle weight, and the results revealed valuable insights.

By changing the bottle material from stainless steel to PET plastic, the results demonstrated that stainless steel 304 is the more desirable option in all three sustainability dimensions. Other possible materials, such as glass, aluminum, and different grades of plastic or steel, can be investigated within this framework to find the most suitable material to meet the design and sustainability requirements. The possibility of changing the materials of different components enables scenario analysis and comparison of different design decision impacts that can assist the designer in improving the product’s holistic sustainability.

Then, after concluding that the bottle material should remain stainless steel, the second step aimed to reduce the bottle weight by changing critical design parameters linked to it. Sensitivity analyses were performed on one variable at a time. The minimum environmental impact in design parameters was found with a 70 mm bottle mouth diameter, but this came at the cost of a significant reduction in economic and social metrics. In addition, a 70 mm bottle diameter resulted in a maximum social impact percentage of 39.30%, which is also an improvement in economic and environmental metrics compared to baseline results. The maximum economic sustainability metric was found at 474,000 USD at a 300 mm bottle height, which has a high social impact and increases the product’s environmental impact. However, the limited variation observed in environmental sustainability across different bottle heights indicates that bottle height does not significantly contribute to the overall environmental impact of the bottle.

The usage as a product attribute plays a significant role in determining the social sustainability metric, which weights 0.3. Usage is a function of other product attributes parameters such as heating efficiency, portability, and washability. The heating efficiency and portability of the bottle are also essential factors that determine the quantity of the product in terms of economic sustainability. Thus, in this example problem, the social and economic sustainability metrics follow a similar pattern in the sensitivity analysis results. In addition, the increased quantity of reusable water bottles demanded generates more jobs, which is another factor that links economic and social sustainability. This shows the essential role played by product attributes associated with the bottle’s performance in socio-economic dimensions.

The change propagation and sustainability assessment methods integrated into this framework distinguish it from the conventional approaches such as LCA, DFE, and eco-design tools that often focus on isolated sustainability dimensions. The results obtained from this analysis align with prior LCA-based studies by highlighting the material selection and end-of-life scenarios as critical impact factors. However, the HSD framework extends these findings by incorporating design interdependencies and exploring sustainability trade-offs during the product design phase. The sensitivity analysis within this framework enables the designers to evaluate the cascading effect of design changes across environmental, economic, and social dimensions, which is lacking in many traditional sustainability tools.

5.2. Case Study Implications on Design

Given the baseline LCA results, it is clear that the reusable water bottle should be designed with an end-of-life scenario in mind, as this life cycle phase was the largest source of environmental impacts. While stainless steel is a recyclable material, this analysis set the system boundaries with a cradle-to-grave perspective; given these results, designers would likely want to explore a cradle-to-cradle approach. Stainless steel also serves as a primary parameter satisfying the product’s heating efficiency and durability requirements, which are linked to the usage and quantity of the product.

The diameter, mouth diameter, and height of the bottle were selected as design decisions that can be changed to improve the overall sustainability metrics of the product effectively. By implementing the MDM and quantifying the dependencies, the analysis could model change propagation and the sensitivity of the design parameters within the proposed framework. The sensitivity results of these three design parameters associated with the bottle captured valuable insights about each sustainability metric that can be the source of improvement and change. These insights serve as a foundation for advancement and fundamental measures to promote positive change and improvement in the product’s sustainability.

The results of the sustainability improvement phase captured the underlying challenges of sustainable design, where a design decision with positive environmental impacts does not translate to a sustainable decision. This is due to compromises in social and economic sustainability metrics. Acknowledging this sustainability dilemma and providing a platform where designers can quantitatively measure each sustainability dimension and perform trade-off analysis based on the organization’s goals is essential. The automated feature of the HSD framework, which uses a custom Python programming language, provides an easy-to-use and flexible sustainable design tool to perform these trade-off analyses. The next challenge arises while simultaneously balancing these three sustainability indicators, where further investigation is required. Previous research has found that an optimal solution between environmental and economic performances could be found by investigating the Pareto or non-inferior surface [64].

5.3. Holistic Sustainable Design Framework

The comprehensive understanding of the underlying structure and interdependencies between design decisions and sustainability metrics revealed within this holistic framework is essential for future research, product designers, and the worldwide effort toward sustainability through product design. This framework constitutes a novel improvement, as it exceeds the conventional fragmented techniques by recognizing the interconnected and mutual dependence among various sustainability metrics and the design decisions of products in different domains. This framework provides crucial opportunities for integrating other methods, such as multidisciplinary design optimization, to achieve sustainability by simultaneously considering the three pillars of sustainability. Through exploring a complex interaction between design decisions and sustainability metrics, valuable understandings can be achieved toward advancements in sustainable design.

Moreover, designers can make informed decisions by quantitatively illustrating the potential consequences of each design parameter’s impact on environmental, social, and economic sustainability indicators. By utilizing this framework in an automated way, the designers are encouraged to explore alternative design possibilities and unlock the potential for changes toward more sustainable products. The insights revealed through this holistic approach could guide designers toward novel solutions that provide environmental sustainability and improve societal well-being and economic viability.

While the reusable water bottle case study demonstrates the framework’s application, applying it to more complex products presents challenges due to increased interdependencies and data requirements. However, the HSD framework’s structured approach can still provide valuable insights for complex systems through mapping change propagation and sustainability trade-offs. The scaling of HSD requires careful data integration and computational resources. It is a valuable tool to reveal the underlying interconnected dependencies across a complex product system and effectively balance environmental, economic, and social factors in product development.

In the broader context, the significance of this HSD framework extends beyond research and product design. It reflects the urgent need for critical global demand to confront sustainability challenges through responsible and mindful product design. The integration of sustainability within product design principles is crucial, considering the growing environmental, social, and economic challenges faced today. The framework proposed in this article offers a structured and systematic approach to tackling these challenges by providing the necessary knowledge to contribute effectively to the broader sustainable development goals.

5.4. Limitations and Future Research Opportunities

Some challenges and limitations arise from the implementation of the HSD framework. First, this framework can be applied to products that already exist or have been completely specified at a detailed design level, where the designers can generate sufficient information to enable the change propagation model. Therefore, at this point, the HSD framework is primarily for the late stages of the new product design process or redesign activities. The manual mapping and quantifying the design parameters is a time-consuming, challenging, but essential task requiring cross-functional design teams to produce a complete and accurate MDM for a complex product. The data collection process is considered a significant challenge in this framework, specifically in capturing and quantifying design dependencies, defining product performance parameters, and assessing sustainability impacts at the unit process level. The complexity associated with this process introduces potential biases in sustainability weighting, where subjective assessments or incomplete data can affect the outcomes. The subjective weighting of sustainability metrics is highly context-dependent and can vary across industries and regulatory environments, resulting in inconsistencies. Developing standardized weighting criteria and adaptive decision-support tools could improve the framework’s application across different engineering domains. The computational demands of the framework increase with product complexity, requiring substantial processing power for evaluating interdependencies and sustainability impacts. Future improvements should explore automated DSM generation and high-performance computing to improve efficiency and scalability.

In addition, a matrix-based method is implemented for LCA, which is simplified and does not account for the inputs and outputs allocation of processes within the production system. The problem of the multi-functionality of unit processes within LCA is where a unit process provides more than one function, requiring an extra allocation step [61]. The simplified method is applied here to reduce the complexity and computational power required for the framework and achieve a feasible solution for the complex sustainability problem. It is also because the comparative LCA results are more important than the absolute results for design decision-making. The uncertainty in LCA has been discussed by scholars [41], and the advancements in this field can be applied to the framework to achieve a more accurate and reliable result for environmental impacts. Moreover, the functional unit defined in the LCA has been identified as one of the essential parts of the analysis that plays a significant role in understanding the product’s environmental impact.

The next steps for this line of research include considering a more comprehensive range of impacts within social and economic sustainability and increasing the accuracy of environmental sustainability results using uncertainty analysis. In addition, clustering algorithms can help to unfold details concerning design parameters’ dependencies. The expectation is to apply this framework to complex products with multiple sub-assemblies and components to examine its application in a challenging situation. Ultimately, it is also possible to implement design optimization within the HSD framework to find the optimal values for each design parameter.

6. Conclusions

This article establishes a novel holistic sustainable design framework capable of demonstrating and quantifying the interdependencies between a product’s design parameters and its environmental, social, and economic sustainability outcomes. A reusable water bottle case study was examined to demonstrate the framework’s application, and the results show promising design improvement insights that can be achieved toward increasing the product’s sustainability. The system definition phase defines the product requirements and system boundaries to map the design decision dependencies across component, product, and sustainability domains using a MDM. Then, in change propagation, the mapped dependencies are quantified to reveal the product’s environmental, social, and economic sustainability impacts. In addition, different design options are investigated in the sustainability improvement phase, and their impact is explored by performing sensitivity analyses. In the case of reusable water bottles, bottle diameter, mouth diameter, and height are changed to find the most effective way of improving the design toward sustainability. The results provided a comprehensive and quantitative understanding of complex underlying dependencies between the design decisions of the product and each dimension of sustainability. The proposed holistic framework enables trade-off analysis and design parameter exploration to support sustainable decision-making in product design. Implementing this method in the design phases of product development can lead to meaningful sustainability improvements by enabling designers to balance lower environmental impacts and improved socio-economic impacts for stakeholders. This framework improves decision-making by quantifying interdependencies between design parameters and sustainability outcomes from the product design perspective. Enabling scenario analysis and sensitivity assessments equips engineers and product managers with a practical tool for optimizing sustainability without compromising functionality or market viability. Future work can explore its scalability to more complex systems and further integration with automation and optimization techniques.

Beyond product improvements, the HSD framework aligns with global sustainability efforts such as SDGs. The framework supports organizations in reducing their environmental footprint while ensuring economic viability and social responsibility by addressing key aspects such as responsible consumption and production (SDG 12), industry innovation and infrastructure (SDG 9), and climate action (SDG 13). Its application can aid in the transition toward circular economy models, where product lifecycle considerations drive design choices that minimize waste and resource depletion.