Introduction

Human activities have a significant impact on the climate and environment on earth, as highlighted in the sixth assessment report of the Intergovernmental Panel on Climate Change (IPCC) [1]. It is stressed that sustainable practices must be introduced on a full scale before it is too late. Such practices are actively pursued in the industry by providing better utilisation of energy and materials and by streamlining manufacturing processes [2, 3]. If successfully implemented, the practices could lead to decoupling economic growth from environmental impact [4], thus allowing manufacturers to continue to prosper and add value to society while lowering their environmental impact [5].

In the pursuit of becoming more environmentally benign, activities such as reuse, repair, and remanufacturing can provide positive effects [6]. By remanufacturing, used, broken, or discarded products – known as cores – can regain like-new performance and condition [7]. Through the industrial process remanufacturing implies, energy and material consumption tend to be reduced compared to outputting an equal-to-new product [8,9,10], allowing to lower the environmental footprint and increase the lucrativeness [11]. Despite the benefits, the remanufacturing industry relative to manufacturing is not widespread. In 2015, the European Remanufacturing Network estimated the European Union’s remanufacturing industry at 1.9% of the manufacturing industry [12], thus a significant growth potential. Similarly, this potential is also apparent in the United States of America [13].

For manufacturing companies to utilise the growth potential, they must realise the value remanufacturing can bring as well as understand how remanufacturing could be initiated. As remanufacturing is a circularity practice, common frameworks could be applied to understand the basic principles of business models and resource flows from a value perspective [14,15,16]. From that point forward, the remanufacturing specifics could be investigated to understand a company’s readiness for remanufacturing in terms of the compatibility of products, the attractiveness of offerings, production systems preparedness, and business model innovation [17]. Additionally, the prerequisites to create a remanufacturing system could be investigated, that is, facilitating the resource and information flows into a remanufacturing process through core acquisition and sales activities [18, 19]. This transition that the initiation implies requires a systematic procedure; otherwise, it is unlikely to reach considerable positive economic and environmental outcomes [20]. In reaching the benefits, the scientific literature provides frameworks to support structuring the initiation [e.g., 11, 21, 22]. Additionally, there is an extensive number of publications describing how to assess the effectiveness of remanufacturing and answer a range of challenges related to remanufacturing, for example, collaboration, competition, customer behaviour, inventory management, location selection, product cannibalisation, production planning, and stakeholder analysis [cf. 23,24,25,26,27].

However, when initiating remanufacturing, the focus is on understanding the economic performance given that the building blocks for a remanufacturing system are experimented with to find a valid solution rather than attempting to reach a high-performance solution for one case. The models that fulfil these requirements are unfortunately lacking in the literature [15, 28, 29] and are not standardised [30]; hence, integrative and practical solutions for remanufacturing [31] cannot be provided to practitioners. As such, it is proposed in this paper that the growth of the remanufacturing industry can be stimulated if the models particularly useful for implementation in industry are identified and categorised. These models should be transparent, simple, and robust to capture the most important factors in assessing the economic performance of initiating remanufacturing. In addition, these models could provide the information needed to motivate industries in approaching and assessing remanufacturing. Hence, the aim of this paper is to identify how remanufacturing initiations can be economically assessed with models particularly suitable for industry. To fulfil the aim, a literature review of scientific publications was conducted. The literature review presents examples of models suitable for supporting industries in initiating remanufacturing and, additionally, it highlights the lack of support for remanufacturing initiation assessments to stimulate the interest in the remanufacturing initiation research area, thus contributing to the transition towards more circular industrial processes and a more sustainable society.

The paper is structured as follows: First, previous research on the economic assessment of remanufacturing is presented (“Previous research on models for the economic assessment of remanufacturing” section). Thereafter, the research approach is described (“Literature review research approach” section), followed by a synthesis of the relevant papers of this literature review and a proposed categorisation framework of the papers (“The two-dimensional framework of data and computer aid requirements” section). Finally, conclusions are presented, and areas for future research are highlighted (“Concluding remarks” section).

Previous research on models for the economic assessment of remanufacturing

Over the years, remanufacturing has been approached diversely. Here, a few selected review studies within the areas of remanufacturing and the closed-loop supply chain (CLSC) have been summarised to show how the call for further research has evolved and identify proposed research gaps related to remanufacturing initiations. These studies were selected to provide a view of remanufacturing research from different time spans. In the next section, the focus is on the findings of the conducted literature review.

Two decades ago, in an early review conducted by Fleischmann et al. [32] on quantitative models for reverse logistics, remanufacturing was described as one of several circularity strategies that retain the value of products. They highlighted remanufacturing as a more demanding strategy than manufacturing since each core requires individual and customised reprocessing to reach expected quality outputs. This variation from one core to another implies that the coordination activities between actors and the planning of material and resource flows became more challenging than in manufacturing modes. Hence, a need to provide research more relevant to remanufacturing contexts appeared and was motivated by a need for further research to integrate forward and reverse supply chains. In relation to this, the existing models for remanufacturing and other circularity strategies were considered data-heavy, thus implying that the models could be challenging to apply for support in practice due to the difficulty in acquiring the needed data. Additionally, the scientific contributions within reverse logistics appeared as not yet mature and fragmented into isolated areas, thus emphasising the call for further research.

A decade later, Guide and Van Wassenhove [31] described a research area where researchers, to a higher degree, used a common language and added it to a common body of knowledge. As such, researchers were no longer targeting fragmented, isolated problems. Guide and Van Wassenhove [31] encapsulated this progress of the research area by viewing 15 years of CLSC research through the lens of business perspective and personal experiences and they draw the conclusion that the research area had gone from infancy to a more mature stage. But there were still gaps to fill in the literature. By focusing on a process flow perspective, they listed three questions in realising remanufacturing, which we interpret as a label for remanufacturing initiations. One was: “Can value be recovered from returns at a reasonable cost (remanufacturing)?” This pinpoints the first piece of a puzzle where the understanding is gained regarding the importance of contemplating whether companies should take steps towards remanufacturing and how these could be supported by economic assessment models. However, it is a puzzle of many pieces yet to be laid, as Guide and Van Wassenhove [31] recommended future research to focus on challenges that are both relevant to industry and academia, as industries need easily applied academic solutions with an industrial language and accessible data. Such models should build upon the body of knowledge and integrate areas to provide practical rather than complex solutions to theoretical problems.

More recent reviews have changed the focus from plotting the research area’s evolution towards understanding the current state of the art by identifying tools and methods related to feasibility assessments of remanufacturing and decision-making; examples are Goodall et al. [24] and Rizova et al. [27]. The former visualised a much higher focus on the three pillars of sustainability than previously, thereby providing a higher level of systems-thinking related to remanufacturing. This indicates that the research area has matured to a state where multiple levels of perspectives are more pronounced, resulting in a possibility to achieve a well-defined understanding of remanufacturing. Another perspective was classifying of decision-making into three management levels: strategic, tactical, and operational. The strategic level shaped questions regarding the long-term view of remanufacturing and the feasibility of initiating remanufacturing, but the contribution within this area was not extensive. The latter, Rizova et al. [27], continued on the same decision-making classification with a wide and comprehensive review of all kinds of remanufacturing decision-making models. One discussed topic was engagements in remanufacturing, that is, topics related to remanufacturing initiations. While it was possible to identify four assessment types – mathematical, analytical, simulation, and conceptual and descriptive – the identified contributions in the remanufacturing initiation area were scarce, especially when excluding products and components for the automobile industry. Conclusively, both reviews, Goodall et al. [24] and Rizova et al. [27], stated a need for assessment models to help companies decide whether remanufacturing initiations are economically sound. It was suggested that such models be practical rather than theoretical to make them suitable for industrial applications.

These literature reviews [24, 27, 31, 32] have shown that the remanufacturing research area has evolved from a fragmented state, where researchers work in silos, to the creation of a common body of knowledge and research focus, and have even shown areas related to remanufacturing initiations. However, despite the call for assessment models suitable for industrial applications being pronounced as early as 1997 in the review by Fleischmann et al. [32], the research community has yet to articulate remanufacturing initiation models that are well-established in industry. In an attempt to showcase this research gap and also contribute to showing examples of models that have successfully been used, we have conducted the literature review presented in this paper. But before going into the research approach and the findings, we would like to take a step back and discuss remanufacturing assessment models without clear relevance to remanufacturing initiations. This is to communicate the distinction between these types of models.

For this discussion, we are relying on the aforementioned literature reviews, as well as on papers excluded in the screening process, that is, papers that propose remanufacturing models for assessing economic performance but do not consider initiation of any kind. These models are still of value for remanufacturing initiations, as they can provide an understanding of how models for the remanufacturing initiation could be built up to consider a systems view of a specific solution. To structure the literature and show in which context they are useful, we rely on the remanufacturing system framework for facilitating remanufacturing, proposed by Barquet et al. [18], to create a go-to source of exemplified economic assessment models. The framework elements are commercialisation of the remanufactured product, design for remanufacturing, employees’ knowledge and skills in remanufacturing, information flow in the remanufacturing system, remanufacturing operation, and reverse supply chain; see Table 1.

Table 1 Identified economic assessment papers and their relation to the remanufacturing system
Full size table

The literature shows that there is a tendency for researchers to place a high focus on models that explore optimal decisions when investigating remanufacturing, and researchers are therefore advancing theoretical modelling techniques to gain managerial insights on expected behaviours in a system. While these are important for the future prosperity and understanding of the remanufacturing area, the models could be difficult to apply for practitioners when they explore the economic value of initiating remanufacturing. Hence, in order to apply the models during remanufacturing initiations, there is a significant effort in adjusting them to new circumstances [77]. With this basis, researchers have continuously highlighted remanufacturing as a complex and uncertain process to master for companies [e.g., 24, 27, 31, 32]. The complexity is inhered from the use of cores as inputs to a remanufacturing process [78] due to their varied conditions [79], uncertain arrival rates [80], and need for custom-tailored disassembly, reprocessing, and reassembly steps based on core-specific conditions [81]. These circumstances result in a distinction between manufacturing and remanufacturing practices, as the latter tend to require a higher level of flexibility. Furthermore, when adding the management aspect of these challenges to the lack of support for assessing the economic performance of initiating remanufacturing [15, 28,29,30], further uncertainties regarding the potential profitability and feasibility are brought to concerns. Moreover, one company is not like another, and thus the required support varies. Therefore, the need for further research on supporting companies in initiating remanufacturing is high. With the present review, we would like to pinpoint a specific research gap and exemplify how the profitability of remanufacturing can be assessed for companies initiating remanufacturing. Such research has also been requested by, for example, Bressanelli et al. [28], Bocken et al. [15], and van Loon and Van Wassenhove [29].

Literature review research approach

In this paper, a systematic literature review was conducted by following the review guidelines of supply chain management, as stated by Durach et al. [82]. The steps consisted of (1) defining the scope and aim of the research, (2) identifying the expectations and filter criteria, (3) determining the search procedure, and, finally, conducting the (4) filtering and (5) synthesis of the literature. Additionally, the works of Rowley and Slack [83] and Schryen et al. [84] were used as further guidance.

The literature review targeted scientific publications applying models to assess the economic performance of initiating remanufacturing suitable for use by practitioners. In this research area, many different terms are used to label the initiation phase and industry implementation of models, for example, adoption, implementation, industrialisation, initiation, and practice. The challenge in using these terms is their broad meaning with uses in many different circumstances irrelevant for remanufacturing initiations. This meant that the searches using these terms were broad, and thus many identified papers lack a unified and expected theme, that is, the economic performance of remanufacturing initiations. To overcome this challenge, three other terms were selected: econom*, profit*, and transition* related to remanufactur* and refurbish*. These search terms are also general, thus providing a broad diversity of papers, but the focus is clearer towards a unified theme related to the research aim. Therefore, these search terms were broad enough to identify the expected targeted sample while simultaneously limiting the number of papers to manageable amounts. With the use of specific filter criteria, the search results were narrowed down to fulfil the specific aim of the paper (See “Literature screening” section).

The choice of search terms evolved during the reviewing process based on preliminary findings, which motivated multiple stages of searches and syntheses. This procedure enabled a flexible approach where the search terms based on preliminary findings of preceding searches could be adjusted. In total, three searches in title, abstract, and keywords were conducted with different key terms and logical operators, as shown in Table 2. The terms econom* and profit* were chosen to target studies that both handled economic and profitability aspects, while transition* was chosen as it is commonly used when referring to a circular economy, where companies transition from linear to circular resource flows. Examples are seen in literature where remanufacturing is indirectly considered, such as, in general terms [85], as a transition to product-service systems [86] or a circular economy [87].

Table 2 Number of papers and unique papers found in the literature databases and searches, respectively
Full size table

The literature databases Scopus, Emerald Insight, Business Source Premier, Web of Science, and Science Direct were selected to cover both engineering and economic research areas within the field of remanufacturing. Only peer-reviewed samples written in English were considered as either journal publications or conference proceedings. The search covered all years available in the literature databases up to August 2021, with the earliest sample identified from 1969. However, as seen in Fig. 1, most of the samples were from more recent years. This resulted in 964 papers that were subject to review.

Fig. 1
figure 1

Distribution of the number of sample papers over time

Full size image

Literature screening

During the screening process, two sets of criteria were derived based on a previously conducted case study on an initiation of remanufacturing for an original equipment manufacturer (OEM) of premium electrical and electronic equipment [19, 22]. If the sampled study could have been applicable during the case study, it was considered relevant to the study’s aim. The first set of criteria was applied during screening on the title and abstract of all 964 papers. These were: (i) Does the paper consider remanufacturing? (ii) Is an approach for assessing the economic performance presented? (iii) Is there a conclusion where remanufacturing from a profitability perspective is considered? Or does the paper investigate an initiation of remanufacturing with consideration to profitability? If the answers were yes for all criteria, the paper was read in full.

For further screening, three additional criteria were applied: (iv) Is it likely that the required data can be collected before industrialising a remanufacturing process? (v) Does the approach provide economic results that pinpoint the benefits of initiating a remanufacturing process? And finally, (vi) Has the approach been demonstrated through practical use in conducting industrial case studies? Given these criteria, twelve papers were identified as relevant; see Table 3.

Table 3 The scientific research papers in alphabetical order with corresponding aim and publication
Full size table

No additional filters were applied when searching the literature databases, meaning that the search process captured some redundancies. This ensured that no samples were left out due to filter criteria bias [100]. As shown in Fig. 2, there were some building and energy-related journals linked to refurbishment that were irrelevant. However, as most targeted journals were perceived as relevant areas for this review, the selection of search terms was considered accurate.

Fig. 2
figure 2

The 20 most common journals of the 964 sampled scientific research papers. Red bars with stripes indicate irrelevant journals, while black bars indicate relevant

Full size image

The two-dimensional framework of data and computer aid requirements

The purpose of this section is to show how remanufacturing initiations can be assessed from a profitability perspective with models particularly suitable for use in industry. To show this, we first defined a basic framework to structure the models, followed by describing how each model approaches the assessment of remanufacturing initiations. Lastly, the synthesis provides a mapping of how these models, in particular, are suitable for this purpose. Using this mapping, a set of recommendations are proposed to guide further research in how to structure economic assessment models for remanufacturing initiations.

Based on the literature review findings, four levels have been proposed and described. These correspond to level one, two, three, and four assessment types, as seen in Fig. 3. The fourth level has not been defined for reasons to be discussed later (“The main characteristics for assessing the economic performance of the remanufacturing initiation” section). The levels are linked to two sets of requirements valuable when assessing remanufacturing initiations, namely, computer aid and data requirements. The data requirements are linked to the accessibility of the needed data. High requirements on data availability indicate that the models require parameters (information) that could be difficult to estimate in the remanufacturing initiation stage. For instance, probability density functions, demand functions, or quality distributions of cores that rely on historical data or specific expertise for accurate estimations are parameters that are perceived as commonly inaccessible, thus hindering the applicability of a model.

Fig. 3
figure 3

Two-dimensional framework of data and computer aid requirements for applying economic assessment models. It consists of four assessment types: levels one, two, three, and four

Full size image

Similarly, the computer aid requirement focuses on how much a given model relies on support from a software application for it to be set up and applied in practice. This requirement is an interpretation for ease of use by assuming that high software requirements increase the threshold for the willingness to use a certain model. While advanced models relying on software support can provide excellent outputs, the threshold is seen as a hindrance for companies not using similar approaches in their daily operations. For example, when a model requires a specific computer software application or algorithm that requires special competence, fewer practitioners have the capability to use it, thus increasing the perceived complexity. In circularity transitions, fully understanding the contributions of offerings, such as remanufacturing, is critical for companies to approve the continuation of a project [101]. As such, confidence in the datasets, models, and software applications could mean the difference between willingly assessing whether a remanufacturing initiation could be a good business opportunity.

The level one assessment type

The level one is distinguished based on its low data and computer aid requirements. Such models use, to a high degree, quantitative reasoning alongside calculations easily performed with pen and paper. The category is proposed as especially suitable for companies that can only collect low amounts of product data and do not heavily rely on assessment models for their daily operations. The models could, to a certain degree, be limited in their capabilities but excel in providing swift results and initial understandings.

One example of a model on this level is the research study by Mont et al. [93]. When they investigated whether a pram manufacturer should lease their products and perform remanufacturing at the end-of-use phase, they first applied a quantitative approach to set the prerequisites for initiating remanufacturing. Thereafter, they assessed the economic performance given the basis that had been created. The assessment focused on mapping the cash flows the new business model induced and assigned minor efforts into the details of remanufacturing. As such, their approach primarily provided insights into how companies on a qualitative level can create a systems view of the new business opportunities, their relation to remanufacturing, and when to apply assessment models as a level of support.

The other research studies within the level one requirement, Lahrour et al. [92], van Loon et al. [96], and van Loon et al. [97], had a higher focus on describing the details of the applied assessment models and how they were used in a given case. Their respective cases were the economic value of remanufacturing air compressors and electric motors, baby strollers, and washing machines. The former had a small company perspective and described how such companies could compare the sales and economic viability of new and remanufactured products, while the two latter had a similar focus as Mont et al. [93] through offering and assessing remanufacturing as part of a leasing scheme. Through these different perspectives, these examples posit the value of applying different approaches based on what information is required to make well-informed decisions.

In explaining how such insights for decision-making are created, the structure of each research study is presented here. To create their systems view, Mont et al. [93], thoroughly explained all steps regarding how the decision-making process was conducted. Part of this was to cover the redesign of the products for ease of remanufacturing by, for example, integrating replaceability of fabrics and other principles to avoid destructive disassembly and shorten the remanufacturing time. For the economic performance, there was a reliance on estimating leasing-related costs, such as additional transports, product reprocessing (i.e., inspection, reconditioning, disassembly), administrative systems, and interests. Furthermore, the reasonings for every decision were described together with the benefits and drawbacks of the explored scenarios, thus facilitating guidelines for initiating remanufacturing given similar business models and needs.

The economic performance is frequently set in relation to alternatives, and this was also the case in Mont et al. [93]. For the air compressor and electric motor case [92], a straightforward comparison of new and remanufactured products was conducted to assess the economic performance. While the strategic dimension and detail in building up the remanufacturing context were not of great detail here, the assessment calculations were presented alongside an approach to involve an inspection process in the analysis to further deepen the understanding. As such, they proposed an approach that could provide practitioners with a quick overview of remanufacturing based on inspecting and determining the value of cores. Moreover, the economic assessment provided detailed insights into how the parameters – such as costs related to core acquisition, transports, labour, and new components – interrelate, meaning that practitioners could gain an understanding of how to develop the foundation of a remanufacturing system.

A higher emphasis on details of more advanced assessment calculations for comparing one-off sales with leasing was provided by van Loon et al. [96]. Despite many mathematical reasonings, similar to Mont et al. [93], the context was built upon a solid foundation, meaning they were straightforward and descriptive for the case at hand. Each step of the calculations has been carefully explained, leading the reader from ideation to the final product. However, from a practitioner’s perspective, collecting the needed data for the assessment model to be used in a similar context is more demanding compared to Mont et al. [93] and Lahrour et al. [92]. However, this is not a negative attribute as it provides a more complete picture of modelling the system of a circular business model with remanufacturing.

Another assessment variant based on the net present value and cost of ownership was proposed by van Loon et al. [97] for the washing machine case. In the leasing-based business model, there were three product quality classes (premium, economy, and basic) that a product could pass through after undergoing remanufacturing. For example, at the end of leasing a new premium product, the product is remanufactured and then leased again as a lower-quality class at a reduced leasing fee. The business model is described in great detail and developed based on scenario creation. Such an approach is beneficial for remanufacturing initiations as it helps to prioritise alternatives, develop long-term strategies, and analyse cases of extremum. Additionally, the case study provides guidelines on setting up leasing-based business models with multiple instances of remanufacturing, contributing with qualitative and quantitative value for remanufacturing initiations.

The level two assessment type

When increasing the requirements on the support from software applications to efficiently apply assessment models for economic performance, the level two is entered. This level is for assessment models that do not have high data requirements, as in level one, but require software support to output the results. The advantage here is that other types of models could be provided, for example, simulation-based models, which provide advantages in scenario creation, with capabilities to utilise large amounts of easily accessible data, or finding optimal approaches given a set of constraints. As such, level two is more capable than level one, but only given that the right type of software support is available.

One example of a simulation model is the publication of Alamerew and Brissaud [88], who approached the remanufacturing of electric vehicle batteries by applying systems dynamics. In their research study, an overview of remanufacturing in a circular economy context was theoretically developed to create a foundation for the logical relations between all circular flows as well as map the case-specific circumstances for their product. This was beneficial since the case involved five companies with various scopes which required a systems view of resource, energy, and information flows to simulate them accurately. The level-one scenario-based approaches, such as the one by van Loon et al. [97], cannot cope with this complexity level due to the high number of flows, giving level two an advantage despite higher requirements. Furthermore, Alamerew and Brissaud’s [88] system dynamics model showed how a simulation-based model could provide critical insights and increase the understanding of assessing remanufacturing initiations, not only from the perspective of providing a cost–benefit analysis but also to understand interconnections within a system and create an understanding of how remanufacturing relates to other activities.

A different approach was taken by Yoda et al. [99] and Alkhayyal [89] by proposing integer programming models to maximise the profitability of a remanufacturing system. While these studies had a lower emphasis on describing the systems perspective, that is, the story behind the reasoning for each step of an initiation, than the aforementioned studies, it was better on presenting and validating a specific solution. As such, for other remanufacturing initiations, the provided value is primarily on how integer programming is used given certain circumstances. For example, Yoda et al. [99] focused on disassembly and assembly strategies for laptops by creating a bill of material (BOM) and listing precedence relationships of detailed component data. This made it possible to further investigate the product structures and measures at the component level. Comparably, the perspective of Alkhayyal [89], for an air conditioning unit, was beyond the BOM focus on a reverse supply chain consisting of multiple collection centres, remanufacturing facilities, and retailers. The broad view extended the lack of sales and coordination activities of Yoda et al. [99], which is important in remanufacturing initiations, to a broad view of creating a potential remanufacturing system. This perspective is valuable for remanufacturing initiations since, even without applying mathematics, the broadness provides an asset in understanding what is required to develop all activities within a remanufacturing system. In that sense, the benefits are twofold: providing an analytical model and increasing the know-how.

The level three assessment type

Level three has the highest requirements on both computer aid and data requirements. As such, the models on this level are comparably more advanced and thus more demanding to apply in remanufacturing initiations. However, this does not mean that the models are less useful because they are more demanding, as their suitability depends on the circumstances specific to a certain initiation. Here, six different approaches are described: linear programming [95], multi-attribute utility theory [90], time-varying remanufacturing strategy [91], heuristic-based analytical tool [94], and decision-support tool with a multi-objective solution algorithm [98]. The main distinction between level three and the lower levels is the more frequent use of heuristics or algorithms to identify valid solutions to the proposed models. From a practitioner’s perspective, this results in less transparency in how each part of the remanufacturing system is interconnected, as a higher level of expertise is required to dissect the relations. Hence, insights into the potential economic performance might be clear from the solutions, but the capabilities to derive informed decisions could be diminished, at least from a general perspective.

Identified here is also a methodological overlap between the level three and two, as linear programming is described on both levels. The case study on automotive parts by van Loon and Van Wassenhove [95] utilised this method. Despite this, the characteristics of the levels are different since the need for accessing high-quality data is higher in level three. In this example, it is due to the higher need for detailed information to accurately estimate the demand for remanufactured products, the remanufacturing yield, and the costs of individual remanufacturing reprocessing steps.

Moreover, the need for accurate and detailed data is profound throughout the identified level-three research studies. The study by Bansal et al. [90] is one example, with its perspective on how the remanufacturing of automotive parts could be viewed in relation to the manufacturing of new products. With the assumption that remanufactured and new products are equivalent, there is no difference in either condition or pricing, thus making it possible to provide one hybrid equation for both manufacturing and remanufacturing. This perspective is unique, as all the other studies consider a difference between new and remanufactured products and the processes needed to output these. As such, this study provides a valuable perspective for remanufacturing initiations, not only with the proposed model but also by providing insights into how to adapt other models given these assumptions. Another unique view was presented by Kwak and Kim [91] with a model that considers the physical and technological obsolescence of returned cores by integrating time-varying capabilities. This principle is valid because the attractiveness of products on the market is not constant. As such, assessing the economic performance of initiating remanufacturing without considerations of time variability, misinterpretations of the potential profitability are likely. Hence, the model provides a valuable perspective and valuable insights that could be included in remanufacturing initiations, even when this model is not suitable for certain circumstances. Naturally, time-varying considerations rely on high-quality data and accurate forecasting models to support the decision of whether to remanufacture or not in comparison to manufacturing new products.

Another perspective for remanufacturing initiations is to relate remanufacturing to alternative value-retention processes, as discussed in the level one model by Lahrour et al. [92]. In relation to this, Yang et al. [98] presented a decision-making framework for end-of-life strategy planning of desktop phones. In terms of needed computer aid when solving, this model ranks as the highest of the identified papers, meaning that practitioners must be well-prepared and knowledgeable in the area. This is mainly due to enforcing a specific solving algorithm – the non-dominated sorting generic algorithm-II (NSGA-II) – to find solutions to a multi-objective problem. However, it also provides a perspective broader than an economic assessment since it has integrated decision-making as well as solutions to assess environmental performance. For remanufacturing initiators prioritising the economic aspect, there are possibilities of relaxing the environmental constraints, making it a purely economic model, and potentially lessening the NSGA-II requirement. A company that initiates remanufacturing or other value-retention processes could use the decision support tool to incrementally investigate if there is any potential to, for example, retain the value of the product or its parts.

Moreover, the manufacturing or remanufacturing of mobile phones and selling them together with two different data plan offerings over two periods was modelled by Ovchinnikov et al. [94]. The model is straightforward, with parameters on an aggregated level, and is either solved for a global optimum or by applying a heuristic. Additionally, the model also covers how to estimate the demand function that is a required input. This is one of the main advantages of the model, as it provides the tools needed to use the model while also providing valuable inputs to support the use of other types of models, such as the ones proposed by Kwak and Kim [91], Bansal et al. [90], and van Loon and Van Wassenhove [95]. One should note that a demand function to fully utilise the models could be difficult to derive without access to historical data on past sales of new or remanufactured products. Further data constraints are related to qualitative considerations not being well-embraced nor transparent in some of the studies, for example, in Ovchinnikov [94] and Bansal et al. [90]. Thus, remanufacturing initiators need to rely on other studies to gain support and insights when determining model parameters.

The main characteristics for assessing the economic performance of the remanufacturing initiation

In the three levels where assessment models were found, the most commonly used assessment techniques were net present value, linear programming, cost of ownership, and various comparisons and simulations of the relationship between manufacturing and remanufacturing; see Table 4. In some cases, multiple techniques were used to gain a wide perspective of the remanufacturing system performance. The scopes of the assessment models varied, as did the data types, their details, and their accuracy. The level one requirement models handled the data demands through the extensive description of business cases, elaborations of the thought processes, and reasonings between scenarios and alternatives [cf. 92, 93, 96, 97]. As such, the proposed models were created based on the data that was accessible. Data that was difficult to acquire was replaced by informed reasoning, scenarios, or well-defined assumptions. As the level requirements increased, the case descriptions became less pronounced and were replaced by more strict assessment models, strict in the sense that they aim to solve a predefined problem rather than describe the milieu where they are most relevant. Which approach is most suitable in remanufacturing initiation situations likely depends on a case-by-case basis of industry circumstances and practices. Certain industries benefit more from a higher level of qualitative descriptions and the related models, while others are more compatible with more quantitative ones; in this selection, the requirement levels could indicate where to investigate first. Based on these needs, six perspectives to frame models to assess the economic performance of remanufacturing were derived, namely:

  1. 1

    Business models – Whether the initiation of remanufacturing is more viable through other arrangements of business activities [e.g., 93, 97].

  2. 2

    Company size – Needs of companies differ due to, for example, different organisational structures, capabilities, and inertia from idea to decision and industrialisation [e.g., 88, 92].

  3. 3

    Relation between new and remanufactured products – Viewing remanufacturing and manufacturing as competitive activities or hybrid complements [e.g., 90 , 91].

  4. 4

    Remanufacturing prerequisites – Identifying the building blocks to create and assess a remanufacturing system [e.g., 93, 96].

  5. 5

    Scenario building – Expanding the base case to identify the best practices and avoid the worst cases [e.g., 88, 97].

  6. 6

    Systems perspective and boundaries – Acknowledging various perspectives of remanufacturing, for example, through the circular economy (high) [88], remanufacturing system (mid) [95], or individual remanufacturing process steps (low) [99].

Table 4 The scientific research papers characterised by model type, product type, case study, and assessment type level
Full size table

Nonetheless, the assessments on the lower levels are considered closer to the aim of this paper since they have lower requirements to be used by practitioners. This analogy is clear from the filtering criteria, as studies not clearly indicating their practitioner orientation or industrialisation capabilities were not considered; that is, highly relevant models were perceived as easier to apply in practice in an industrial context. This leads to the discussion of the level four requirement of Fig. 3, comprised of proposed models of low computer aid and high data requirements, which did not contain any of the identified studies. The lack of level four is intuitive from the perspective that the more easily accessible and interpretable the datasets, the lower the need for support from integrated software applications. As such, the perceived tendency is a correlation between high data requirements and the need for computer support, meaning that level four does not fulfil a need.

Furthermore, remanufacturing initiations could also be assessed from other perspectives, such as the environmental. While this is not a part of the research aim, this perspective is highlighted and considered in the assessments of some of the research studies; see, for example, [89, 91, 94, 95, 98]. Incorporating both economic and environmental considerations in a remanufacturing initiation is beneficial from a sustainability perspective and is, therefore, what the industry should strive for. The reason this dual perspective was not presented in this literature review is due to the assumption that remanufacturing is always more environmentally friendly than manufacturing. This is often the case thanks to its energy- and material-retention capabilities [8,9,10,11], which are achieved by taking back cores at suitable points in time and by not overconsuming resources. As such, companies can instead focus on their economic performance and the value of transforming their business strategies.

Concluding remarks

There are many researchers developing economic assessments for remanufacturing. Within this study, a high number of reviewed papers, 964, and an identified sample of twelve papers indicate that there are few research studies on assessing the economic performance of remanufacturing initiations, at least from the perspective of being particularly suitable for industry. The results highlight the need for further research in the area to support companies that have products suitable for remanufacturing but have not yet taken the step towards an initiation due to an inability to assess their profitability. This is a clear research gap, not only because there is a lack of such models as shown in this literature review, but because there are multiple calls explicitly advocated by several researchers, for example, Bressanelli et al. [28], Bocken et al. [15], and van Loon and Van Wassenhove [29], and it has been indicated in extensive reviews through various perspectives related to remanufacturing systems and planning; see [24, 27, 31, 32]. With this basis, we would like to highlight the research gap further and structure the call for more research in the remanufacturing initiation area. The research on the economic performance for remanufacturing initiations should be practitioner-oriented to ease the suitability of assessment methods to be used in industry by developing transparent, simple, and robust models.

In achieving this characteristic, it is first necessary to acknowledge that products and industries are unique [14, 102] and thus have different needs. It is, therefore, not possible to identify a single solution for all circumstances. In this literature review, we have highlighted economic assessment models ranging from comparing business models [e.g., 88, 93, 96] and the trade-off between remanufacturing and manufacturing activities [e.g., 91, 92, 97] to creating several scenarios within singles cases [e.g., 88, 97]. This range of models cannot be seen as substitutes; rather, they are complements in creating a comprehensive understanding. In a general setting, we indicate six perspectives to frame the scope of these models: (1) business models, (2) company size, (3) relation between new and remanufactured products, (4) remanufacturing prerequisites, (5) scenario building, and (6) systems perspective and boundaries. These perspectives are not definitive but are shown here to highlight the value of developing further contributions for remanufacturing initiations of various types and scopes to enable the identification of suitable assessment models for companies.

Furthermore, to show the difficulty level of applying the economic assessments for practitioners in initiating remanufacturing, the identified assessment models were categorised into a two-dimensional framework indicating data and computer aid requirements. Four assessment types are proposed: levels one, two, three, and four. The first and the third have low and high-end requirements for both dimensions, respectively, while level two (four) is positioned at low (high) data and high (low) computer aid requirements; see Fig. 3. There were no assessment models identified for level four, and it is reasoned that the level does not fulfil a need for support from integrated software applications since the datasets are easily accessible and interpretable. Level one was perceived as easy to use in remanufacturing initiations but could lack the features and capabilities of more complex models. Analogously, the level two and three requirement levels are more demanding. The low levels tend to provide detailed qualitative case descriptions before defining and applying an assessment model. As such, these research publications are much more informative in the sense that they provide motivation for the contextual benefits of remanufacturing and why the used assessment approach is valid [e.g., 88, 93, 96]. Remanufacturing initiators can, therefore, take inspiration from the case descriptions to gain an understanding of best practices for their own unique cases independently of the provided models. Taking inspiration from the higher levels to build business strategies and create broad perspectives of remanufacturing systems is more difficult as the focus is directed towards solving pre-defined problems and then defining the logical flow of a proposed model. However, there is no one-size-fits-all solution for all industries and products, as indicated by the different scopes of the models, meaning that depending on the given product and company circumstances, their suitability and effectiveness vary [14, 102]. As such, certain companies benefit more from low-level models while others more from a high level. The variation further emphasises the need for more research within the area, as a higher number of developed models would increase the probability that a suitable model can be found and used during certain circumstances without needing to develop a new one from sketch. While the suitability of the models is case dependent, it is suggested that remanufacturing initiators view the three requirement levels as a recommended order to search for an appropriate model, that is, starting with level one models and then continuing with level two and three until a match is found, since the qualitative contributions are seen as valuable.

Moreover, most of the studies that were considered irrelevant to the aim of the review have proposed models whose practical industrial applicability or suitability for remanufacturing initiations have not been shown, meaning that it is difficult to anticipate how well they perform when applied in practice for remanufacturing initiations. However, these models could still be useful for remanufacturing initiations as they can aid in adapting models by providing an understanding of how certain parts of a remanufacturing system could be modelled. To ease the use of supporting models for remanufacturing initiations, Table 1 provides a selection related to all areas of the remanufacturing system [cf. 18]. Hence, practitioners can customise assessment models for their own needs. Nonetheless, studies of and models for remanufacturing initiations provide a more streamlined approach for remanufacturing initiators to use, and therefore it is emphasised to focus further research on conceptualising assessment models and describing initiation cases in practice.

Furthermore, due to high complexity, some models were interpreted as difficult to apply and solve. As such, the perceived complexity of a model could be a hindering factor for practitioners. This is especially the case for smaller companies that have challenges in recruiting the top of the class employees. Moreover, complex models tend to require complex data, such as statistical density functions of both sales of remanufactured products and acquisition of cores; see [33, 37, 68]. Such inputs were interpreted as inaccessible for companies initiating remanufacturing. Moreover, data regarding the customer’s willingness to pay for remanufactured products and the environmental consciousness of these customers are other examples, as are extensive data on demand rates, quality of cores, cannibalisation, and remanufacturing yield. Behaviour data and data on product variations are more likely to be available at mature companies that already have an active and developed remanufacturing process. However, such companies have a lower need for the models that this review attempted to identify than companies with low remanufacturing experience, as their operations are already well developed, and their internal processes could provide indications of what improvements are needed. The companies in a higher need of modelling techniques are those that have not been operating for a long period of time or that have never experienced remanufacturing before. These companies will likely prefer transparent, simple, and robust models designed with a practitioner-orientation in mind without requirements on over-detailed data on the required operations.

In theory, the missing required data could be estimated. However, the difficulty of estimating parameters for remanufacturers of varied matureness has been acknowledged in the literature [64]. Companies initiating remanufacturing would therefore need to rely on uncertain estimations to use the models. Moreover, even complex models tend to simplify by introducing assumptions, thus leading to inaccuracies [103]. Estimations and assumptions in combination could lead to higher uncertainty and misleading model outputs due to inaccurate numbers. The difficulty of handling the inaccuracy increases in tandem with the complexity of the models. Thus, if many uncertain parameters are estimated, there is a risk that the model output and the actual case will be far apart [104].

As a final point, this study also has its limits. Despite applying a systematic procedure to ensure objectivity throughout the reviewing process, there are cases of subjective perception due to manual content analysis. The reviewing process has mainly been conducted by a single author, meaning that a second opinion was not utilised when screening. Therefore, the identified papers should be seen as examples of papers relevant to the purpose rather than a finite and decisive sample. The significant contribution lies in the trend of the low number of identified papers, which signifies a need for further research.