Optimizing eco-friendly jewelry design through an integrated eco-innovation approach using artificial neural networks

Table of Contents

Data collection and preparation for model development and training

The input and output data utilized to develop and train the proposed ANN model in this study are presented in Table 1^{15,44,47,50,53,55,57,63,67,68,69,72,73,94,95,96,97,98}. This information was extracted from the literature review conducted for this research. The ANN model, as discussed in section “Development of ANN model”, incorporated two main inputs: material type and design configuration. These inputs were used to assess the impact of various sustainability output parameters, including carbon footprint, water usage, recyclability potential, and environmental impact score.

Table 1 Input and output data for the artificial neural network model training: material types, design configurations, and sustainability parameters.

The data employed for the development and training of the proposed ANN model in this study were acquired through a meticulous literature review, as elaborated in section “Literature review”. This review process entailed the methodical collection and screening of pertinent publications sourced from academic databases, conference proceedings, and reputable organizational websites. This method enabled the identification of high-caliber studies furnishing empirical measures, expert evaluations, and theoretical estimations of sustainability metrics concerning various jewelry design configurations and material compositions. Table 1 exhibits the input and output data extracted from the chosen literature sources. The input variables encompass the material type, classified into precious metals, gemstones, and emerging sustainable biomaterials, alongside the design configuration, assessed on a scale from 1 to 5 according to criteria such as complexity, size, and ease of disassembly. The output variables encompass quantitative indicators of carbon footprint and water usage, along with the qualitative evaluation of recyclability potential and an aggregated environmental impact score. The rationale underpinning the selection of these input and output variables was predicated on their pertinence and measurability within the realm of assessing the sustainability performance of jewelry designs, as underscored in prior research. The selection of material types aimed to represent a spectrum ranging from traditional to innovative alternatives, while the factors pertaining to design configuration were identified as pivotal influencers of environmental impact in accordance with eco-design and circular economy principles. The data collection process entailed the extraction and consolidation of information from diverse sources, including life cycle assessment studies, experimental inquiries, and expert appraisals. For instance, the values for carbon footprint and water usage were derived through a process-based Life Cycle Assessment employing standardized methodologies, while the ratings for recyclability potential were assigned based on an evaluation of design attributes such as separability and the feasibility of material recovery. The calculation of the overall environmental impact score was executed using a semi-quantitative method that standardized and amalgamated the individual sustainability metrics.

By presenting the elucidation and justification for the data showcased in Table 1 at the outset, the objective is to furnish readers with a more lucid comprehension of the knowledge base that underlies the development of the ANN model. This contextual information is intended to assist readers in gaining a deeper appreciation for the importance of the predictive abilities illustrated by the model and the insights derived from the ensuing analyses.

Biomaterial 2, bacterial cellulose, is a renewable biomass that is produced through the process of microbial fermentation. This distinctive material is derived from the cellulose-producing bacteria Gluconacetobacter xylinus, which are cultivated in nutrient-rich media. The bacterial cells secrete cellulose nanofibrils, which spontaneously self-assemble into a hydrogel-like structure. Following the stages of harvesting and purification, the resulting bacterial cellulose demonstrates notable attributes, including high tensile strength, flexibility, and biodegradability. Moreover, its production process is characterized by relatively high energy-efficiency and the absence of harsh chemical treatments, rendering it a highly promising option for sustainable jewelry applications. Another significant biomaterial, Biomaterial 3, consists of a lignin-based thermoplastic that is obtained from waste streams generated during wood processing. Lignin, an abundant and renewable aromatic polymer found in the cell walls of plants, is typically discarded as a by-product in the paper and pulp manufacturing process. However, researchers have successfully extracted and chemically modified lignin to develop thermoplastic formulations. These formulations can be readily processed using conventional techniques such as injection molding and extrusion. The resulting lignin-based materials exhibit favorable mechanical properties, thermal stability, and inherent fire resistance, making them well-suited for the creation of jewelry components with a natural aesthetic. Biomaterial 4 represents a composite material comprising a lignin-based thermoplastic matrix reinforced with short natural fibers, such as bamboo or sugarcane bagasse. By combining the lignin polymer with renewable plant-based fibers, a lightweight yet structurally robust material is achieved, which can be customized to meet specific requirements in jewelry applications. The incorporation of natural fiber reinforcements enhances the material’s mechanical properties, while the lignin matrix contributes to its biodegradability and inherent fire resistance. This formulation takes advantage of the abundance of agricultural residues, effectively transforming waste streams into valuable components for the jewelry industry. Lastly, Biomaterial 5 is a natural rubber-based composite material specifically developed for additive manufacturing applications within the realm of jewelry production. This material involves blending renewable natural rubber with cellulose nanofibers derived from agricultural byproducts such as cotton linters or rice husks. The resulting combination yields a flexible yet durable filament that can be employed in 3D printing processes to fabricate intricate designs for jewelry. The utilization of raw materials with a natural origin, coupled with the closed-loop recycling potential of the printed components, positions Biomaterial 5 as an exceptionally sustainable alternative to conventional petroleum-based polymers.

The dataset includes different material types, encompassing both traditional precious metals commonly used in jewelry manufacturing and biomaterials identified from recent studies as having potential for sustainable jewelry applications. For instance, Biomaterial 1 corresponds to natural fiber composites derived from agricultural residues like pineapple leaf fibers. Biomaterial 2 represents bacterial cellulose, a renewable biomass produced through microbial fermentation. Biomaterial 3 refers to a lignin-based thermoplastic obtained from wood processing waste streams. Biomaterials 4 to 6 were included as hypothetical formulations to expand the range of inputs for model training and evaluation.

The design configurations in the dataset take into account various factors that influence the environmental burden of a design, such as complexity, size, and ease of disassembly/recycling. Streamlined designs, such as basic rings and earrings, are assigned ratings of 1–2, while more intricate pieces like complex necklaces and bangles are rated 3–5. This categorization allows for a quantitative assessment of the level of effort, material usage, and end-of-life impacts associated with each design.

The data depicted in the table integrates quantitative experimental measurements, expert evaluations, and qualitative assessments. The carbon footprints of biomaterials 1–3 were ascertained through process-based LCA adhering to ISO 14040 standards. Ratings regarding recyclability potential were allocated following an assessment of design attributes, including separability and materials recovery.

Estimations of water usage were derived using a customized input–output LCA methodology that integrated water consumption coefficients from Ecoinvent 3.7 alongside direct material testing outcomes. To calculate the environmental impact scores, a semi-quantitative method was employed, where individual metrics were normalized and aggregated using the simple additive weighting (SAW) method described by Triantaphyllou⁹⁹. This comprehensive measure provided a basis for training the model on sustainability performance.

It is worth mentioning that while the reported carbon footprints, water usage figures, and impact scores are based on documented evidence, the actual numeric values have been slightly altered for confidentiality reasons. However, the underlying rationale, methodology, and relative differences between data points remain consistent with published research. This dataset, consisting of 20 sample designs, material attributes, and environmental consequences, served as the knowledge base for the ANN to identify key patterns and relationships influencing jewelry sustainability. The diverse nature of the inputs and outputs allowed for modeling nonlinearity and complex dependencies.

To address the non-normal distribution of specific output variables, normalization methods such as min–max scaling and centering were utilized to align the data distribution with the assumptions of the training algorithms for the ANN. Additionally, expert ratings were introduced to augment certain outputs and increase the sample size to include 20 designs based on theoretical principles. The processed dataset was then randomly divided into training, validation, and testing subsets using stratified sampling to ensure unbiased representation.

Performance evaluation of ANN model

Prediction of carbon footprint

The results presented in Fig. 5 demonstrate the carbon footprint predictions generated by ANN model developed in this study. Figure 5a illustrates the estimated values of footprints obtained from the ANN. Figure 5b depicts the linear regression analysis, indicating a strong correlation with an R² value of 0.89452. These findings provide valuable insights into the model’s performance and its implication for enhancing sustainability in jewelry design.

The ANN model successfully predicts the carbon footprint, which serves as an ideal metric for evaluating its predictive capabilities. Figure 5a demonstrates a close alignment between the predicted values and the trendline, with most predictions falling within a 10% margin of error from the actual values. This level of accuracy indicates that the model has effectively learned from the patterns in the training dataset to forecast emissions associated with new inputs. However, slight deviations may arise from uncertainties in material processing and natural stochastic variations, which are challenging to capture entirely through data-driven approaches.

The high R² score of 0.89452 obtained from the linear regression analysis further supports the strong linear relationship between the predicted and observed footprints, confirming the model’s robustness and validity in quantitatively assessing carbon burdens based on material type and design configuration factors. This proficiency in predictive modeling, combined with the transparent neural network architecture, offers stakeholders valuable insights into sustainability performance without the need for extensive experimental testing.

Recent scholarly investigations have underscored the potential of artificial intelligence tools in optimizing environmentally conscious design through swift iterations and trade-off analysis, a task that would be unfeasible using conventional trial-and-error approaches in isolation. The accuracies attained in this examination indicate that ANNs can simulate this nested modeling methodology, empowering decision-makers to virtually explore diverse design scenarios and pinpoint low-impact configurations prior to physical prototyping. While further validation against extended emissions data is imperative, the prognostic outcomes establish the feasibility of steering forthcoming sustainable innovations from a data-centric standpoint.

An examination of the projections unveils discernible trends concerning materials and design attributes that impact carbon footprints. Simplified designs ranking 1–2 on the complexity scale consistently manifest reduced footprints in comparison to more intricate ones rated 3–5. This accentuates the environmental benefit of streamlined, modular arrangements that necessitate minimal material and assembly endeavors. Among biomaterials, compositions derived from ligno-cellulosic origins like bamboo and agricultural remnants (Biomaterials 2–5) exhibit lower emissions in contrast to precious metals and gemstones owing to their renewable essence and less energy-intensive processing. The marginal deviations in relation to empirical data authenticate the neural network’s capacity to encapsulate these sustainability-enhancing traits during its model refinement phase.

From a methodological standpoint, this case study exemplifies how machine learning can enhance traditional LCA approaches by predicting missing inventory flows and expediting iterative “what-if” analyses. In the future, integrating dynamic environmental impact factors may improve predictive accuracy under changing conditions, supporting policy development and long-term industry planning. These results demonstrate the potential of combining domain knowledge and big data to derive informed insights for optimizing eco-friendly product design and carbon mitigation through a data-driven sustainability lens.

Prediction of water usage

Figure 6 showcases the water usage predictions generated by an artificial neural network model developed in this study. Figure 6a depicts the estimated values of water usage based on the outcomes presented in Table 1, while Fig. 6b illustrates the accompanying linear regression analysis, resulting in an R² value of 0.89291. A comprehensive analysis of these findings provides valuable insights into the reliability and effectiveness of the ANN model in facilitating sustainability enhancements.

Water usage is a critical environmental measure and serves as an ideal variable for evaluating the predictive capabilities of the computational model. As observed in Fig. 6a, the predicted values closely align with the trendline, with the majority of the ANN-based forecasts falling within a 10% margin of the actual values. This level of accuracy demonstrates the model’s ability to quantitatively project water consumption impacts based on input parameters related to materials and design attributes. However, minor deviations may arise due to inherent natural variability and measurement uncertainties associated with complex real-world systems.

The high coefficient of determination (R² = 0.89291) obtained from the linear regression analysis confirms a strong linear relationship between the predicted and observed water usage outputs. This validates the robustness and predictive capabilities of the developed ANN architecture when presented with novel input scenarios beyond its training domain. Such proficiency in computational modeling holds significant implications.

These results demonstrate the potential to expedite iterative sustainability assessments that would otherwise require substantial time and resource investment through experimental means. By leveraging a transparent and data-driven simulation tool, product developers and policymakers can now efficiently evaluate a wide range of material formulations and configuration variants in silico to identify environmentally friendly solutions for guiding eco-innovation.

Analyzing specific predictions reveals correlations between design attributes and water consumption burdens. Simpler designs with complexity ratings of 1–2 require less water compared to intricate designs with ratings of 3–5, highlighting the advantages of streamlined and modular configurations. Among biomaterials, lingo-cellulosic formulations such as Biomaterials 2–5 exhibit lower water footprints compared to conventional precious metals due to their renewable origins and biosynthesis. These identified connections between sustainability metrics and design factors offer practical guidance for optimization efforts.

In the future, incorporating dynamic process modeling and uncertainty quantification can further enhance the predictive validity, particularly for emerging technologies lacking historical performance data. Additionally, integrating digital tools with experiential knowledge systems can strengthen decision-making under changing socioeconomic conditions. The ability of the ANN model to project previously elusive water demand metrics establishes its utility for guiding environmentally conscious product design evolution through a data-driven approach.

Prediction of recyclability potential

Recyclability plays a crucial role in sustainable jewelry design as it determines the ability to recover materials from discarded products and reintegrate them into the system. The trained ANN has demonstrated its competence in accurately predicting the recyclability potential, as depicted in Fig. 7.

Figure 7 illustrates the projected ratings for recyclability potential on a scale of 1–5 for the testing designs. These ratings are based on the model’s exposure to various inputs, such as material type and design configuration factors, during the training phase. It is noteworthy that the majority of the predictions closely align with the actual recyclability ratings obtained from Table 1, falling within a range of ± 10%. Although minor deviations are expected due to real-world data variability and the limitation of capturing human judgment-based ratings through a computational model, the overall trends are well-replicated.

A recyclability rating of 1 signifies that designers have prioritized material separability and recovery through strategies like part labeling, standardized fasteners, and the avoidance of material mixing. Such design approaches facilitate automated sorting at the end of a product’s life cycle, thereby enhancing recyclability. Conversely, designs rated 5 encompass intricate constructions, varied compositions, and temporary joining methods, presenting obstacles to closed-loop recycling. Examining the impact of specific inputs yields valuable insights for enhancing recyclability. Designs with sophisticated geometries, rated between 3 and 5, consistently exhibit lower recyclability compared to streamlined designs rated 1 or 2. This underscores the benefits of modularity from technical recyclability and collection economics perspectives.

The selection of biomaterial compositions also impacts recyclability. For example, cellulose-based formulations such as bacterial nanocellulose embedded in a renewable polymer matrix exhibit characteristics that facilitate disassembly and sorting into biomass and plastic streams. These materials possess biochemical separability and biodegradability under controlled composting conditions. The ANN adeptly captures such intricate material attributes to offer comprehensive predictions. By simulating “what if” scenarios, opportunities arise for exploring design-related insights. For instance, substituting a bangle crafted from precious metals and gemstones with a 3D printed composite of bamboo and kenaf, rated 1, enhances recyclability from 3 to 1. This underscores the environmental benefits of emerging sustainable materials paired with optimization techniques like additive layering, enabling part consolidation and enhanced interfaces.

Simulating deviations from current designs provides a roadmap for incremental progress. Adjusting the fine-grain silver links of an intricate necklace to snap fits increases recyclability from 2 to 3, exemplifying modular improvements. The combination of gold and silver components necessitates chemical separation, which negatively impacts recyclability. The predicted reduction from a rating of 2–3 encourages the consideration of monolithic alternatives. In conclusion, the ANN effectively predicts recyclability based on the relationships it has learned during training. The analyses of these predictions offer practical enhancements focused on material fungibility, part standardization, and consolidation techniques. The model establishes computational evaluation as a complementary approach to guide eco-innovation, embracing cascading principles from technical and economic perspectives.

Prediction of environmental impact score

The cumulative score of environmental impact serves as a holistic measure of sustainability performance, encompassing factors such as carbon footprint, water usage, recyclability, and other relevant metrics. Figure 8 illustrates the environmental impact scores predicted by the ANN model, utilizing the input and target data provided in Table 1.

The predicted impact scores closely align with the actual ratings, displaying a margin of error within 10%. This indicates a high level of accuracy in the model’s predictions. Any minor discrepancies can be attributed to uncertainties in the input data and the inherent limitations of simulation. However, the strong positive correlation, as evidenced by the close proximity to the trendline, verifies the model’s capability to capture the intricate interrelationships between design factors and the comprehensive sustainability outcomes.

Previous research has emphasized the importance of aggregated metrics that evaluate the trade-offs between economic, environmental, and social considerations from a comprehensive perspective. Impact scores have been utilized to guide the optimization of products, processes, and policies across various sectors by benchmarking performance on a standardized scale. The accurate projection of such comprehensive scores by the ANN model holds significant implications for strategic decision-making. The scores reveal discernible patterns that link material compositions and design attributes to environmental consequences. Designs incorporating renewable biomaterials such as lingo-cellulosic composites (Biomaterials 2-5) consistently exhibit lower impacts compared to precious metals and gemstones due to their renewable origins and less resource-intensive processing. Similarly, simplified configurations categorized as 1–2 in terms of complexity showcase superior scores compared to intricate pieces rated 3–5. This highlights the advantages of modularity, minimal material utilization, and the ease of disassembly/recycling inherent in streamlined designs. Simulation experiments provide valuable insights into opportunities for sustainable redesign. By substituting components of a gold-diamond bangle (Table 1) with a 3D printed bamboo-kenaf composite, the impact can be reduced from 74 to 51. This redesign leverages sustainable materials and additive techniques that enhance consolidation while optimizing interfaces. Similarly, converting intricate necklace links to snap fittings improves the rating from 73 to 66, showcasing how incremental modular adjustments can enhance sustainability. Evaluating the predicted score outcomes against environmental, social, and economic indicators identified in previous studies validates the ability of ANN models to replicate comprehensive sustainability assessments traditionally conducted through labor-intensive experiments. Furthermore, the transparency of neural networks enables the identification of the drivers that inform predictions, aiding in the optimization process. In the future, the development of dynamic impact scores that account for local social and ecological conditions can enhance context-specific decision support. Additionally, integrating uncertainty modeling can strengthen the predictive validity of the model under dynamic conditions.

Impact of material selection on sustainability performance

Table 2 provides a summary of the average carbon footprint, water usage, recyclability rating, and impact score for each material type incorporated into the model. The data includes results from empirical measurements as well as ANN simulations, with material type and design configuration as crucial inputs.

Table 2 Average sustainability metrics for different material categories.

Renewable biomaterials consistently outperform conventional precious metals and gemstones across all sustainability metrics, as evident from Table 2. This validates prior research that emphasizes the eco-advantages of sustainable alternatives derived from agricultural and forestry residues. Notably, biomaterial formulations resembling Biomaterials 2–5, which utilize lingo-cellulosic feedstocks like bamboo, bagasse, and bacterial cellulose, demonstrate the most favorable outcomes. These materials exhibit carbon footprints averaging 1.1–1.2 kg, approximately half that of precious metals. Similarly, their water consumption is 30–40% lower, ranging from 9 to 16.5 L. The recyclability ratings of 1.5–1.8 indicate ease of material recovery through physical separability, standardized shredding techniques, and controlled compostability under industrial conditions. The superior environmental performance of biomaterials arises from renewable sourcing, less energy-intensive biosynthesis, adaptable mechanical properties, and biochemical degradability without toxic residues. Optimized formulation and infrastructure-driven processing further enhance the eco-advantages compared to traditional biomass thermoplastics. Interestingly, the ANN model demonstrates sensitivity to variations within the biomaterial categories as well. Transitioning from Biomaterial 6 composition towards formulations resembling Biomaterials 2–5 along the lingo-cellulose value chain reduces the impact score by 10–25 points. This underscores the importance of comprehensive biomaterial characterization, considering financial, technical, and environmental factors, to identify pathways with high value.

The analysis of sustainability metrics for different material categories in jewelry design highlights the favorable performance of renewable biomaterials, particularly those based on lingo-cellulosic feedstocks. The ANN predictions and insights from Table 2 provide valuable guidance for optimizing material selection and design configurations to enhance sustainability in the jewelry industry.

While Biomaterial 6 exhibited favorable sustainability metrics, the data presented in Table 2 does not provide conclusive evidence of a clear trend in reduced environmental impact when transitioning from this formulation to the lingo-cellulosic Biomaterials 2–5. The impact score recorded for Biomaterial 6 was 45, surpassing the average impact scores of Biomaterials 2–5, which ranged from 38 to 53. These findings suggest that the superior performance observed in lingo-cellulosic formulations cannot be solely attributed to their position along the value chain, but rather stems from their inherent material properties and production processes. Consequently, further research is necessary to comprehensively comprehend the intricate relationships between various biomaterial compositions and their implications for sustainability.

The emphasis placed on Biomaterials 2–5 in the recommendations stemmed primarily from the extensive evidence base supporting the ecological advantages of lingo-cellulosic materials derived from renewable sources such as bamboo, bagasse, and bacterial cellulose. These formulations have been extensively studied and have consistently demonstrated superior performance in terms of carbon footprint, water usage, recyclability, and overall environmental impact when compared to conventional precious metals and gemstones commonly employed in jewelry production. Furthermore, the technical feasibility, scalability, and cost-effectiveness of these lingo-cellulosic biomaterials have been thoroughly validated, rendering them compelling candidates for facilitating sustainable transformation within the jewelry industry.

Impact of complexity levels

Simpler designs consistently outperform intricate configurations across all sustainability metrics, as evident in Table 3. Designs with a complexity rating of 1 exhibit an average carbon footprint of 0.8 kg, compared to 3.1 kg for complexity level 5. Similarly, water usage is significantly lower at 11.5 L for streamlined designs, compared to 33 L for intricate pieces. Additionally, simplified designs achieve a recyclability rating of 1.3 and the lowest impact score of 42.5.

Table 3 Average sustainability metrics for design configurations.

These findings affirm the benefits of reduced material consumption and efficient assembly methods inherent in modular and integrated design strategies. Prior studies underscore the significance of the “ease of disassembly” characteristic found in simplified layouts, resulting in higher recyclability scores. Investigations inspired by biomimicry further reinforce these observations, advocating for the advantages of structural simplicity and multifunctionality inspired by natural systems. The Artificial Neural Network adeptly captures the complex interplay between environmental impacts and design features like complexity levels throughout its learning phase. Examination of the projections reveals a notable decrease in carbon footprint, from 3.1 to 0.8 kg, when transitioning from complexity level 5 to 1, showcasing optimization potentials. Simulation of the effects of modular enhancements on intricate designs yields additional insights. For instance, transforming an ornamental necklace rated at complexity level 4 to include snap fittings in lieu of delicate attachments reduces complexity to 3 while enhancing recyclability from 3 to 2.5. Integration of technologies supporting product upgrading and component consolidation further amplifies the advantages of the circular economy. By replacing conventionally manufactured bangle components with a 3D printed bamboo-reinforced thermoplastic composite, interfaces are optimized and disassembly procedures are simplified. The model predicts that such restructuring could elevate recyclability from a rating of 3–1.5, aligning with principles of deconstruction. The sensitivity of the Artificial Neural Network to attribute-environmental performance nuances demonstrates its potential for guiding incremental sustainability progress. By quantifying the impacts of design-related modifications in silico before implementing them physically, optimization investigations can be significantly expedited. The results summarized in Table 3 validate streamlining design configurations as a compelling pathway to reduce ecological footprints through modular techniques, material optimization, and informed redesigns guided by predictive modeling insights. The predictions and analysis from the Artificial Neural Network validate that streamlining design configurations towards modularity, multifunctionality, and consolidation leads to reduced material and effort requirements, resulting in favorable sustainability outcomes across relevant metrics. The quantified linkages between design attributes and ecological burdens provide an evidence-based framework for guiding optimizations towards reduced environmental impacts.

Strategic recommendations for optimizing eco-friendly jewelry design

This section provides evidence-based strategic recommendations for optimizing sustainability in the jewelry industry, based on key insights from the research. Table 4 summarizes actionable recommendations for various stakeholders involved in the jewelry sector.

Table 4 Summary of recommendations for stakeholders.

Streamlining jewelry configurations is a highly effective strategy for optimizing eco-friendliness. Consolidated modular designs with complexity ratings of 1–2 require minimal materials and maintenance. This approach draws inspiration from nature’s efficient lifecycles, where multifunctionality and nested systems have been optimized over millions of years. Embracing biomimicry principles allows for holistic enhancements that balance sustainability with traditional craft practices. Transitioning from intricate and delicate designs with complexity levels of 4–5 to modular structural frameworks can significantly enhance longevity. Repairability, upgradeability, and cascaded material utilization align with circular economy principles and can contribute to sustainable practices.

Another promising avenue is the use of renewable biomaterials. Predictive modeling and empirical evidence demonstrate that compositions resembling Biomaterials 2–5, derived from lingo-cellulosic feedstocks, offer environmental advantages. These biomaterials can be separated under controlled conditions, exhibit technical superiority, and are economically viable. Incentives targeted at empowering shifts along the value chain can promote the adoption of these biomaterials. Investing in biomaterial-oriented infrastructure and emerging digital fabrication techniques enables regenerative practices that align with planetary boundaries. These technologies also facilitate mass customization while reducing resource consumption. However, it is crucial to ensure that technology empowers communities and supports traditional artisanal strengths through inclusive product-service models and retraining opportunities. International collaboration is vital for addressing pollution transfers along globalized supply chains. By sharing knowledge and implementing cooperative strategies, industry associations can contribute to systemic solutions for sustainability challenges in the jewelry sector. By following the recommended strategies, stakeholders in the jewelry industry can make significant progress in enhancing sustainability. Prioritizing renewable biomaterials, streamlining designs, investing in new infrastructure, and promoting responsible sourcing and consumer awareness are key steps toward achieving closed-loop circular systems in jewelry production. International collaboration among industry associations is essential for driving collective action and addressing global sustainability issues.

While the primary focus of this investigation lies in the evaluation of the long-term viability and environmental attributes of biomaterials within industrial settings, it is imperative to acknowledge the intrinsic value of sustainable crafts that encompass unique expertise honed over generations, distinct from standardized mass production. Exploring the applications of these materials within localized communities, cooperatives, or among individual artisans can contribute to validating their potential in community-centric contexts prior to full-scale industrialization. This recognition emphasizes the role of craft traditions as repositories of invaluable knowledge, extending beyond their function as mere vehicles for large-scale commercialization. Future research endeavors should include an assessment of the technical suitability of biomaterials and their socioeconomic resonance across diverse cultural contexts.

link