1. Introduction
The global shift towards greater sustainability, driven by increased public awareness and stricter government regulations, is forcing companies to operate more sustainably (Reference Kämmler-Burrak and KruckKämmler-Burrak & Kruck, 2023). The European Union (EU) is at the forefront of regulatory advancements, exerting significant pressure on businesses through initiatives such as the Corporate Sustainability Reporting Directive (CSRD) (European Parliament, 2022), the EU Green Deal (Reference DoddDodd, 2023), and the Ecodesign for Sustainable Products Regulation (ESPR). Around 95% of companies believe that sustainability will become even more important over the next five years (Reference Clemens, Fritsche, Herring, Schleyer, Laufenberg, Brückner, Kühmann, Heering, Schwarzkopf and YükselClemens et al., 2022). Future corporate success must thus be intrinsically linked to sustainability, requiring a fundamental shift towards carbon-neutral products and processes (Reference Stern and ValeroStern & Valero, 2021). This transformation is commonly known as decarbonization.
Parallel to this need for decarbonization, industrial production is increasingly digitalized. In Germany, 98% of industrial companies anticipate enhanced efficiency through digital technologies, with 91% already investing in the development of digital factories (Reference Geissbauer, Schrauf, Berttram and CheraghiGeissbauer et al., 2017). Cyber-physical production systems (CPPS), which integrate digital technologies into production, are used by the manufacturing industry for production process automation and control (Reference Wang, Törngren and OnoriWang et al., 2015). CPPS represent a subset of cyber-physical systems (CPS) that combine IT and software components (cyber) with mechanical and electronic parts (physical), which communicate via a data infrastructure (Reference BroyBroy, 2010). By facilitating improvements in material and energy efficiency, flexibility, and reliability (Reference ThiedeThiede, 2018), CPPS are projected to contribute 15-25% to industrial decarbonization efforts (Reference ThiedeThiede, 2022). Therefore, CPPS are at the centre of the transformation of industrial production towards digitalization and sustainability (Reference Kürpick, Kühn, Olszewski and DumitrescuKürpick et al., 2023).
Driven by their anticipated benefits, the number of CPPS is projected to more than double between 2024 and 2032, exceeding 37 billion by 2032 (Transforma Insights, 2024). However, the rapid expansion of CPPS presents a potential environmental challenge. Without adequate environmental life cycle management, the increased production and eventual disposal of these systems could significantly increase industrial emissions, counteracting their intended benefits (Reference Rogall, Mennenga, Herrmann and ThiedeRogall et al., 2022). In a worst-case scenario, interconnected devices are predicted to account for a global carbon footprint of over 1,000 Mt carbon dioxide equivalents per year by 2027 (Reference Pirson and BolPirson & Bol, 2021). The growing deployment of CPPS implies that they will increasingly account for a substantial portion of industrial companies’ carbon emissions. Therefore, it is crucial to minimize the carbon footprint of CPPS across their entire lifecycle, encompassing manufacturing, operations, and end-of-life phases (Reference Schöllhammer, Cortes‐Cornax, Lago, Ramautar, Roncancio, Overbeek and EspañaSchöllhammer et al., 2024).
While many manufacturing companies employ product carbon footprint (PCF) assessments as an initial step and as a key component of their environmental management, 80% of product-related environmental impacts are determined at the product design stage (Reference SuSu, 2020). Hence, there is a need to shift from post-production analysis to ecodesign. Conducting life-cycle-based analyses for CPPS and determining corresponding decarbonization measures is a resource-intensive process (Reference Schreiner, Kürpick, Kühn and DumitrescuSchreiner et al., 2023). This is due to the unique complexities, intricacies and the novelty of CPPS (Reference Cortès Cornax, Lago and RoncancioCortès Cornax et al., 2022). While industrial companies urgently require ecodesign guidance for CPPS, research has yet to provide adequate solutions. A vast amount of ecodesign literature, methods and tools exists, but ecodesign approaches suited for the application to CPPS are lacking.
Based on the described problem statement, the research question of this paper is how to effectively decarbonize CPPS at the design stage. Inspired by the research process of Ulrich et al. Reference Ulrich, Dyllick and Probst(1984) the paper is organized as follows: Based on the identified practical problem, an analysis of problem-relevant theories from basic sciences and methods from formal sciences results in the identification of a research need for a framework for decarbonizing CPPS at the design stage. Based on these specific needs, the framework and its elements are developed. To validate the framework, it is applied to a real-world application example of an intelligent industrial control valve. To conclude, the results are critically reflected and an outlook to further research is provided (Figure 1).
Figure 1. Research process of applied science by Ulrich et al. and structure/results of this paper
2. State of the art
The research field of ecodesign has gained significant momentum over the past decades (Reference Brones and Monteiro de CarvalhoBrones & Monteiro de Carvalho, 2015; Reference Schäfer and LöwerSchäfer & Löwer, 2020). However, there is a clear need for more focussed research on design-for-decarbonization respectively low-carbon design for CPPS. Existing approaches vary significantly in their scope, from simple, measure-oriented checklists to detailed analytical tools. A review of various ecodesign approaches, including those focused on generic ecodesign, ecodesign for production, low-carbon design, and the systematization of sustainability measures, reveals four key deficiencies in current literature.
First, existing approaches inadequately account for CPPS. This inadequacy has two facets: Most approaches can only be applied to CPPS to a limited extent and the decarbonization opportunities created by digital technologies are not considered. To the best of the authors knowledge, Rogall et al. Reference Rogall, Mennenga, Herrmann and Thiede(2022) is the first and only study to address both of these facets by proposing a methodological approach for developing sustainability-oriented CPPS based on cross-linkage, cross-impact and environmental analyses. However, this approach does not provide recommendations for operational design, nor does it consider the entire product life cycle, and thus does not adequately account for different decarbonization pathways. Second, there is a lack of approaches tailored to decarbonization. This especially applies to established generic ecodesign approaches such as the Ecodesign PILOT by Wimmer and Züst Reference Wimmer and Züst(2003), the Life Cycle Design Strategy (LiDS)-Wheel by Brezet and van Hemel Reference Brezet and van Hemel(1997), the design for environment guidelines by Telenko et al. Reference Telenko, O’Rourke, Conner Seepersad and Webber(2016) and the SustainPro tool by Carvalho et al. Reference Carvalho, Matos and Gani(2013). Decarbonization is considered by most approaches, but the inclusion of several other environmental aspects leads to less efficiency in achieving the most pressing environmental challenge of decarbonization (UN, 2015). Third, many solutions are poorly integrated with existing design processes. In particular, low-carbon design (Reference Miklautsch and WoschankMiklautsch & Woschank, 2022; Reference He, Wang, Huang and WangHe et al., 2015; Reference Lu, Zhou, Xiao, Chang and TianLu et al., 2017) and ecodesign for production approaches (Reference Dombrowski, Schmidt and SchmidtchenDombrowski et al., 2014; Reference Züst, Züst, Schudeleit and WegenerZüst et al., 2016) often only consider individual stages of the design process and do not address how to integrate them into industry-wide and company-specific processes and standards. Fourth, most low-carbon and production-specific approaches do not address all product life cycle phases. While life cycle thinking is an established concept in generic ecodesign, this is not the case for approaches in the domain of production and low-carbon design. These approaches often focus on individual life cycle phases or aspects, such as energy use (Reference Züst, Züst, Schudeleit and WegenerZüst et al., 2016), high emission parts of equipment products (Reference Lu, Zhou, Xiao, Chang and TianLu et al., 2017), or transportation (Reference Miklautsch and WoschankMiklautsch & Woschank, 2022). To achieve holistic decarbonization of products, it is essential to consider the entire product life cycle, ensuring that carbon hotspots are identified and optimized on a global scale rather than focusing on localized impacts.
In conclusion, there is a need to extend the current state of the art in addressing the decarbonization challenges posed by the complex nature of CPPS and the integration of physical and digital components. As a result, there are theoretical deficits regarding the effective application of ecodesign for CPPS. Consequently, there is a need for research into the design-for-decarbonization of CPPS including methodological support to enable effective decarbonization at an early stage of product design.
3. Framework for decarbonizing CPPS at the design stage
This section presents the framework for decarbonizing CPPS at the design stage. The framework enables product design teams to achieve data-based, product-specific design decisions by detailing decarbonization strategies into effective measures based on the carbon hotspots along the life cycle. The framework is developed by conducting an integrative literature review. This type of review is chosen because it is particularly suited to synthesizing and conceptualizing knowledge into new frameworks (Reference SnyderSnyder, 2019). The review is conducted based on the guidelines proposed by Torraco Reference Torraco(2005). Literature is searched through Google Scholar, Scopus, econbiz, ProQuest One Business, and backward and forward citation mining. The relevant knowledge base comprises generic ecodesign, ecodesign for production, low-carbon design, and the systematization of sustainability measures (Section 2). Key elements from reviewed approaches are analyzed and interconnected to develop the framework. As a result, the framework consists of three steps: (1) determination of decarbonization (sub-)strategies, (2) identification, and (3) prioritization of decarbonization measures (Figure 2). Methodological support is provided for each step to facilitate the design process: the decarbonization wheel (first step), the decarbonization measure catalogue (second step), and the measure prioritization logic (third step).
Figure 2. Framework for decarbonizing CPPS at the design stage
Based on the PCF and the decarbonization target, the initial step involves formulating a design-for-decarbonization strategy and subsequently developing effective sub-strategies using the decarbonization wheel framework (Section 3.1). In a second step, concrete decarbonization measures are identified based on the prioritized sub-strategies. This involves deriving generic design measures by applying the measure catalogue (Section 3.2). As a result, a longlist of potential measures is created, which are ranked at the operational level in a third step using a measure prioritization logic (Section 3.3). The underlying principle of the framework is to gradually reduce the abstraction level of the design advice on three levels: decarbonization strategies, sub-strategies, and measures (Figure 3), as proposed in similar manner by Pigosso Reference Pigosso, McAloone and Rozenfeld(2014).
Figure 3. Three levels of abstraction - decarbonization strategies, sub-strategies, and measures
Based on the integrative literature review, the framework differentiates between 6 strategies, 26 sub-strategies and 97 measures. To understand its principle, consider the following example: The strategy “Reduced impact during use” can be achieved through the sub-strategy “Lower energy consumption”. This can be accomplished by implementing measures like “Energy consumption dashboard for machine user” or “Demand-oriented operation mode”. This structure supports the operationalization of strategies into concrete decarbonization measures and forms the basis for the following elaborations.
3.1. Determination of decarbonization (sub-)strategies
The initial step involves defining specific decarbonization strategies and sub-strategies. The decarbonization wheel is a helpful concept to provide methodological support and guidance in this step (Figure 4). It is inspired by the work of Brezet and van Hemel Reference Brezet and van Hemel(1997) and adapted to meet the requirements of decarbonizing CPPS. More specifically, the concept utilizes the basic idea and structure of the LiDS-Wheel and adds specific sub-strategies and measures according to the results of the integrated literature review.
Figure 4. Decarbonization wheel including decarbonization strategies and sub-strategies
The five axes of the decarbonization wheel represent the five product life cycle phases: materials, production, distribution, use, and end-of-life (Reference VezzoliVezzoli, 2018). Each life cycle phase is associated with one respective decarbonization strategy except for the use phase, which is associated with two strategies. Each strategy is subdivided into and defined by several sub-strategies. The application of the decarbonization wheel requires five sub-steps in total.
The first sub-step is to input the PCF and the decarbonization target for the product under consideration. It is recommended to use a detailed bottom-up PCF, i.e. calculated based on the bill of materials (BOM) and the bill of process (BOP) based on a cradle-to-grave consideration. The PCF is entered into the decarbonization wheel according to the five life cycle phases. Decarbonization targets include either short- and medium-term reduction targets or long-term climate neutrality targets. These should be set in accordance with established initiatives such as the Science Based Targets initiative (SBTi) and consider the development cycles of the respective product generation.
In a second sub-step, an ABC analysis is conducted to categorize the strategies based on the Pareto principle (Reference Flores and Clay WhybarkFlores & Clay Whybark, 1986; Reference SandersSanders, 1987). As a widely used analytical methodology, the ABC analysis is helpful for classifying and prioritizing elements based on their importance (Reference KaufmannKaufmann, 2025; Reference DickieDickie, 1951). In the context of the decarbonization wheel, A strategies therefore address 80% of emissions, B strategies a further 15% and C strategies the last 5% of emissions. The ABC analysis provides a target-based categorization of the strategies and reveals the strategies that need to be implemented to fulfil the reduction need quantified by the decarbonization target. Ambitious decarbonization targets require a more holistic approach by considering a broader range of strategies. Less ambitious reduction targets can be achieved by implementing a low number of measures, thus the consideration of only one or two strategies may be sufficient. The ambition of a decarbonization target should be evaluated by considering factors such as product characteristics, industry standards, and decarbonization targets set by both the company and relevant government bodies.
The third sub-step is to refine the selected strategies through the prioritization and selection of sub-strategies, as outlined in sub-steps three to five. The sub-strategies are assigned to the strategies by the decarbonization wheel (Figure 4). To assess the relative potential of each sub-strategy, pairwise comparisons are used. In general, the pairwise comparison technique evaluates the relative importance of different alternatives by comparing two items at a time (Reference SaatySaaty, 1980). Here it is used on sub-strategy level to evaluate the potential of each sub-strategy to reduce emissions. Only sub-strategies are considered that are associated with one of the strategies selected in the second sub-step. It is recommended that various pairwise comparisons of members from a total of four dimensions are carried out: technology, environment, economy and system perspective (Reference Viebahn, Schüwer, Holtz, Pastowski, Klingen and AydemirViebahn et al., 2024). To combine the pairwise comparisons from the four dimensions, another pairwise comparison between the dimensions can be conducted. The pairwise comparison yields prioritized sub-strategies.
In a fourth sub-step a second ABC analysis is conducted for classifying the prioritized sub-strategies according to their reduction potential (potential-based categorization). The aim is to ensure a sufficient number of sub-strategies to implement the decarbonization target and to allow a combined assessment in the subsequent step. Thus, the basis for the second ABC analysis are the prioritized sub-strategies as a result of the pairwise comparison. Based on these results, the ABC analysis yields a categorization of the measures according to their potential to substantially reduce total carbon reductions.
The fifth sub-step entails a combined assessment of the sub-strategies to establish a complete ranking. This is done based on the additive combination of the two ABC analyses: The target-based categorization of the strategies (first ABC analysis) and the potential-based categorization of the sub-strategies (second ABC analysis). The target-based categorization refers to the top-down reduction need, while the potential-based categorization relates to the bottom-up reduction potential. As the importance of the strategies must be prioritized over the importance of the sub-strategies, the aggregated types are ranked as follows from most to least important: AA, AB, AC, BA, BB, BC, CA, CB, CC. From this classification the aggregated priority of each sub-strategy is derived. For sub-strategies with the same combined type, their priority rank is ordered by their weights from the pairwise comparison. The result of the combination is a hierarchical prioritization of sub-strategies. Based on this joint consideration and the decarbonization target, the design team must decide which sub-strategies to include and exclude.
3.2. Selection of decarbonization measures
To bridge the gap between the strategic level and its implementation a decarbonization measure catalogue is proposed for the second step of the framework. The catalogue represents a collection of 97 design measures for decarbonization resulting from the integrative literature review. It was developed in three steps: measure collection, measure synthesis, and measure elaboration. The aim of the catalogue is to provide a comprehensive overview of the specific measures that contribute to each design-for-decarbonization sub-strategy (Figure 5). This enables an efficient pre-selection of measures by seamlessly integrating the catalogue into the decarbonization wheel.
Figure 5. Excerpt from the decarbonization measure catalogue
The entries of the measure catalogue indicate the strategies and sub-strategies each measure is attributable to. Each measure is provided with a fact sheet containing a measure description and information on the impact, expenditures, risks, requirements of the measure and an allocation to the three sustainability principles of sufficiency, consistency, and efficiency. Moreover, an indication of involvement of digital technologies and references for further information are provided. This enables the design team to easily derive measures based on the design-for-decarbonization sub-strategy. The full version of the catalogue can be retrieved by contacting the authors. All measures associated with one of the selected sub-strategies are considered. The result is a longlist of decarbonization measures.
3.3. Prioritization of decarbonization measures
Once design-for-decarbonization sub-strategies are established and suitable measures are identified, it is necessary to determine which measures are most relevant in the third step of the framework. For this purpose, a measure prioritization logic is proposed. The prioritization logic is an adaption of the systematization for sustainability measures by Brinken et al. Reference Brinken, Trojahn and Behrendt(2022). The idea of the logic is a two-dimensional systematization of measures in a matrix (Figure 6).
Figure 6. Measure prioritization logic
The horizontal hierarchy constitutes prioritization by the impact mechanism, while the vertical hierarchy follows the impact relevance. The systematization by impact mechanism stipulates a hierarchy by the three archetypal sustainability principles (Reference HuberHuber, 1995): sufficiency, consistency, and efficiency. Of these three principles, sufficiency claims the highest priority, as it addresses the problem at its root and leads to absolute reduction of environmental impact by decoupling economic value from environmental burden and resource use. Consistency is prioritized after sufficiency. The idea of consistency is shifting the frugality of sufficiency towards frugality regarding requirements and functions, by fulfilling requirements through material and energy flows that are aligned with natural capital. The lowest priority is attributed to efficiency. Efficiency aims to improve the ratio between economic output and environmental input (Reference HuberHuber, 1995). However, this results in two shortcomings: Efficiency measures only allow relative improvements with comparatively small impact and often lead to rebound effects.
The impact relevance indicates the importance of a measure based on the previously prioritized sub-strategies. Accordingly, the sub-strategies are to be entered in descending order of importance from top to bottom. As a result, the structuring according to impact mechanism and impact relevance results in a prioritization of the measures with decreasing importance from the top left to the bottom right corner.
4. Validation of the framework
The framework is applied to the real-world application example of an intelligent industrial control valve. The valve is designed for industrial process applications for intelligent control of flow and pressure of steam, neutral gases, vapors, and liquids. The example is abstracted and simplified to respect sensitive information and to fit the scope of this paper. Figure 7 provides an overview of the validation.
Figure 7. Application example of an intelligent industrial control valve
The application example focuses on cradle-to-gate PCF information, i.e. 60% of emissions are associated with the materials used, 10% with upstream transport and 30% with production. The short-term reduction target has been set to 10%. With this input the decarbonization wheel is initialized. As part of the ABC analysis, the “Reduced material impact” strategy associated with materials and the “Optimized production technique” strategy associated with production are categorized as A strategies and the “Optimized distribution system” strategy associated with upstream transport as B strategy. Due to the low share of transport in total emissions and its B categorization, only the sub-strategies associated with the two A strategies are considered further. These can be extracted from the decarbonization wheel. The sub-strategies are subsequently weighted by a pairwise comparison with the head of design, who covers the dimensions of technology and economics, and a sustainability expert, who covers the environmental and systemic dimension. The results are combined in a further pairwise comparison. With the subsequent second ABC analysis the five sub-strategies “Selection of low-impact materials”, “Reduction of materials usage”, “Cleaner energy source”, “Alternative production techniques”, “Reused/remanufactured components” and “Procure low-carbon components” are prioritized.
Based on the prioritized sub-strategies specific measures are derived. For this purpose, the measure catalogue is used, which enables the assignment of specific measures to the respective sub-strategies. For the application example, this results in a longlist of 19 measures to decarbonize the valve. These include for example: “Prefer materials with lower energy content & carbon footprint”, “Introduce digital product passport”, “Use renewable energy resources”, “Use low energy input production techniques”, or “Reuse and refurbish parts and components of sold products”.
To structure this longlist of measures according to impact, the measure prioritization logic is applied. First, the sub-strategies with the highest potential are entered in the vertical dimension of the matrix according to their priority ranking from the second ABC analysis. Subsequently, all measures under consideration are entered in the matrix according to their allocation to the respective sub-strategy and their allocation to the sustainability principle. The allocation to the sustainability principle is given in the respective fact sheet of the measure. If a measure is associated with several sub-strategies, then it is assigned to the sub-strategy with the highest priority. The resulting positioning in the matrix reveals the priority to be given to each measure. Measures positioned towards the top left corner have the highest priority, while measures towards the bottom right corner have the lowest priority. The resulting prioritization of measures should not be interpreted as an ultimate ranking of measures. It is much more intended to provide a first, low-effort systematization which bridges the transition from conceptual design to detailed design and provides a basis for further investigations. Using the prioritization logic, the measures “Prefer materials with lower energy content & carbon footprint” and “Use low energy input production techniques” are prioritized as the most promising for the application example.
Based on this, the measures are transferred to the intelligent control valve, specified, and simulated within the PCF model. As a result of the simulation, by changing the material and production technique of two components, the 10% decarbonization target is achieved at the same manufacturing cost.
5. Conclusion
In conclusion, this paper developed a framework for the decarbonization of CPPS at the design stage. Using the PCF and the decarbonization target as inputs, the underlying principle of the framework is a gradual reduction of the abstraction level of the design advice: from strategies to sub-strategies to measures. This is achieved in three process steps, each supplemented by methodological support. First, the decarbonization wheel is developed as a tool for defining a design-for-decarbonization (sub-) strategy. Second, a catalogue of 97 decarbonization measures is compiled based on an integrative literature review. The catalogue is linked to the 6 strategies and 26 sub-strategies of the decarbonization wheel and enables an efficient selection of measures. Third, a measure prioritization logic enables the prioritization of selected measures based on the sustainability principles of sufficiency, consistency, and efficiency. The scope of the framework reaches from the conceptual design stage of strategy building and ends with an initial list of ranked measures at the operational level. It addresses the entire product life cycle from cradle-to-grave and focusses on carbon emissions for holistic and effective impact reduction. Further, the design advice is tailored to CPPS and includes environmental improvements through digital technologies. The measure-based approach provides product design teams with concrete design options, thereby accelerating the design-for-decarbonization process for CPPS.
However, the framework exhibits limitations. Firstly, the framework does not consider trade-offs and synergies between measures. Further, the impact and costs of measures are not quantified. Quantified values such as the carbon reduction impact and implementation costs of measures, would facilitate decision-making, but are challenging to estimate generically. A potential barrier for applying the framework is the necessity of the PCF of the product under consideration. While the reliability of model outputs fundamentally depends on the quality, accuracy and completeness of input data, many companies do not have this data readily available. In addition, further application and validation of the framework is required. Lastly, the framework does not provide for radical solutions. Finding radical solutions and developing new product concepts requires innovative problem-solving methods and techniques that are incompatible with the measure-based approach sought in this paper.
Based on the results of this paper, further research needs can be identified. The literature review for decarbonization measures must be substantiated and expanded to include grey literature such as sustainability reports. Further research is needed on the generalizability of the framework to increase its applicability and impact to a broader product range. Furthermore, a higher availability of efficient and potentially automated PCF assessment methods, tools, and data for novel and complex products such as CPPS is necessary for increasing the effectiveness of this design-for-decarbonization approach.
Author contributions
The manuscript is based on C.A.’s master’s thesis, operationally supervised by S.N. and strategically supervised by V.T. C.A. wrote the first draft of the manuscript, with S.N. writing the validation and S.N. and V.T. critically refining the content. C.A. developed an initial illustration of the figures, which were fundamentally enhanced by S.N. All authors have critically reviewed and approved the manuscript. All authors read and approved the published version of the manuscript.
Open access
