The Role of Digital Technologies in Operationalizing the Circular Economy Transition: A Systematic Literature Review

: The enabling role of Digital Technologies towards the Circular Economy transition has been recognized. Nonetheless, to support the transition, the operationalization of the discourse is still needed. The present study performs a systematic literature review, deepening the knowledge on the role of Digital Technologies in operationalizing the Circular Economy transition. The analysis is shaped according to the ReSOLVE framework, as it has been recognized as able to operationally guide industrial ﬁrms towards the Circular Economy transition. Despite the broad focus on the topic by the extant literature, the results of the analysis show limited Circular Economy aspects addressed and speciﬁc technologies considered, making it difﬁcult to have a complete overview on the implementation of Digital Technologies in the Circular Economy transition, operatively addressing it. Shortcomings are identiﬁed regarding the lack of an integrated and holistic analysis of the relationships, the need for investigating the decision-making process and speciﬁc Circular Economy practices, all from an empirical perspective. The paper eventually suggests streams for further research while offering theoretical and practical implications.


Introduction
Deep and rapid economic, environmental and social changes are taking place, shaping the political, managerial and academic discourses [1,2].The industry is not exempt from these current macro-trends and opportunities arise for two specific paradigms, namely Circular Economy (CE) and Industry 4.0 (I4.0) [3,4].
CE focuses on closing the material loop, shifting from a linear economy to a circular one, decreasing material extraction, waste disposal and, consequently, environmental pressure [5,6].CE can be applied at different levels, namely micro (single firm, from a single product to the advertisement), meso (industrial systems and networks) and macro (society or country) [6].Although focusing more on an environmental perspective, it is impossible to separate CE from the economy and society, which links CE to the concept of strong sustainability [7,8].On the other hand, I4.0 enables intelligent factories and products, providing opportunities for enhanced performance in terms of production activities, organizational strategies, business models and skills [9,10].A central role in I4.0 is played by Digital Technologies (DTs) [11].
The two concepts have been largely addressed in a separate manner; nonetheless, in the last years, they started being integrated [12].From a general perspective, it is widely accepted that DTs can enable the CE transition [13].DTs indeed allow more efficient and flexible processes [14], while also providing transparent access to product data and resource consumption [13].Despite the growing interest in the role of DTs as an enabler for CE transition, some points remain still not properly addressed.Particularly, focusing on the CE micro level, the need for making the overall discourse more operational, addressing the different phases of the CE transition [15], has been underlined [16].The Regenerate, Share, Optimize, Loop, Virtualize, Exchange (ReSOLVE) framework has been identified as an important tool to operationally guide industrial firms [17]; despite its relevance, only a few studies so far have focused on the enabling role of DTs in the context of the ReSOLVE framework.The majority of the contributions, indeed, still focus only on specific CE aspects, such as recycling or resource efficiency.On the other hand, contributions focusing on the ReSOLVE consider the role of very few and specific DTs.Both situations underlined the lack of an overall, comprehensive and integrated approach towards the investigation of the role of DTs as an enabler for CE.
Based on the considerations above, the present work aims at conducting a systematic literature review, so to better understand the possible role of DTs within the context of the ReSOLVE framework.To the best of the authors' knowledge, such an analysis is still missing, and a detailed identification of the relationships among all the available DTs and the action areas of the ReSOLVE framework needs to be investigated.
The remainder of the paper is structured as follows.We provided a background on the frameworks for the analysis of CE and DTs (Section 2).Following, we described the systematic literature review methodology, clearly outlining the steps (Section 3).After a descriptive evaluation of the results (Section 4), we analyzed the literature in terms of emerging themes, addressing the possible role of DTs in the context of the ReSOLVE framework (Section 5).We then discussed some specific issues for which additional research is necessary (Section 6).Finally (Section 7), we outlined pivotal implications of our study and paved the way for further research.

Materials
The section introduces the frameworks used in the present work for the analysis of the literature, in terms of content for both CE and DTs.As anticipated in the previous section, to understand if and how DTs can enable the CE transition, we focused on the relationship between DTs and the ReSOLVE framework.Particularly, as a limited set of contributions addresses directly the ReSOLVE framework and its different action areas, we decided to further link specific CE aspects to the ReSOLVE areas.

Circular Economy
Despite the soaring relevance of CE in the current debate, a common definition and agreement on pivotal concepts is not easy to find [6,18].Nonetheless, to allow the CE transition in the industrial sector, the concept must be disclosed from a concrete viewpoint; this would support industrial firms to fully exploit resources while maintaining their value and minimizing environmental impact [19].
Among the different frameworks conceptualizing the CE, the discourse has been largely focused on the 3Rs (Reduce, Reuse, Recycle) model [20].The model soon evolved into the 6Rs model (Redesign, Reduce, Reuse, Remanufacture, Recycle, Recover) and then into the 9 (10)Rs model (Refuse, Redesign, Reduce, Reuse, Repair, Refurbish, Remanufacture, Repurpose, Recycle, Recover) [6,21].The Rs or waste hierarchy models have been included in the butterfly diagram [22] proposed by the Ellen MacArthur Foundation [23].Following a cradle-to-cradle approach, the diagram highlights the difference between the loop for biological and technical nutrients.As for the technical loop, activities such as reuse, refurbishment and remanufacturing are strongly recommended [24].Focusing on the need for industrial firms to move from linear to circular modes of production, and particularly on the opportunities deriving by the technical loop [25], the Ellen MacArthur Foundation [26] developed the ReSOLVE framework.The framework entails major circular business opportunities [27].It proposes six areas of actions for implementing the CE transition, namely: Regenerate, Share, Optimize, Loop, Virtualize and Exchange (Figure 1).Adapted from [26,28].
Several strategies can be related to the six areas [27,29], allowing the definition of operational actions for the CE transition [30].A comprehensive and largely shared overview of possible CE operational actions for the CE transition is offered by Rosa et al. [31], who identified 10 aspects: Circular Business Model (CBM), i.e., the overarching concept; Digital Transformation (DIGIT); Disassembly (DISAS); Lifecycle Management (LIFEC); Recycling (RECYC); Remanufacturing (REMAN); Resource Efficiency (RESOU); Reuse (REUSE); Smart Services (SMSER) and Supply Chain Management (SCM).Leveraging on the indications provided by Kalmykova et al. [29] and Lewandowski [27], we linked the 10 CE aspects to the ReSOLVE areas (Table 1).This operation would allow a clear classification of the literature according to the ReSOLVE areas.


Cloud/fog/edge technologies (CLOUD): architectural models enabling pervasive, convenient and on-demand network access to shared resources such as networks or servers [36];  Cybersecurity and blockchain (CYB): technologies, tools, guidelines and policies guaranteeing the protection of the cyber environment, allowing confidentiality, integrity and availability of data [37];  Horizontal/Vertical system integration (HVSYS): universal data integration network, enabling an automated value chain within or among firms by means of linking products, plants, manufacturers, customers and suppliers [38];  Simulation (SIM): a real-time reflection of the physical world (products, machines, human beings) in virtual models; it can allow testing and optimizing systems before implementing the physical change [31];  Augmented reality (AR): technologies providing an interactive computer simulation, immersing the user in a programmed environment, simulating a sense of reality whether in the sight, in the hearing or the tactile sense [39];  Autonomous robots (ROBs): robots able to operate completely autonomously, to interact with each other and to cooperate with human beings; sensors and control units facilitate the autonomous decision-making process and symbiotic work with humans [40];  Additive manufacturing (AM): production of items directly from CAD models, with fabrication performed layering the material; AM offers the valuable ability to build parts with geometrical and material complexity, not feasible with traditional manufacturing processes [41].

Methods
The present study employs a systematic literature review to identify, select and critically appraise relevant research.To guarantee a scientific and replicable approach, we referred to the steps proposed by Tranfield and Denyer [42], proceeding through (i) questions formulation, (ii) source identification, (iii) study selection and evaluation, (iv) analysis and synthesis and (v) reporting and using results.Additionally, to increase clarity and transparency, we used the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)statement as the backbone of our analysis [43].

Question Formulation
We formulated the question according to the CIMO logic [42], combining a problematic context (C), for which the design proposition suggests a certain intervention (I), to produce, through specified generative mechanisms (M), the intended outcome (O) [44].
RQ: How (M) and in which condition (C) DTs (I) can enable the CE transition (O)?

Source Identification
For source identification, we investigated the Scopus database [4,45,46].We performed a keyword-based search, interconnecting keywords deriving from the two paradigms, using terms related to CE (circular economy, circularity) and terms related to DTs (digital*, Industry 4.0, IoT, Internet of Things, Artificial Intelligence, AI) For the latter group of keywords, we selected the most frequent terms used in similar works [47,48]; additionally, we based our choice on the insights provided by Munirathinam [49] and Lee et al. [50], according to whom Internet of Things and Artificial Intelligence could be possibly used as synonymous of Industry 4.0.As for exclusion criteria, we limited the analysis to contributions published in English from the year 2000 onwards.We thus performed the following query: (TITLE-ABS-KEY ("circular economy") OR TITLE-ABS-KEY ("circularity")) AND (TITLE-ABS-KEY ("digital*") OR TITLE-ABS-KEY ("industry 4.0") OR TITLE-ABS-KEY ("iot") OR TITLE-ABS-KEY ("internet of things") OR TITLE-ABS-KEY ("artificial intelligence") OR TITLE-ABS-KEY ("ai")) AND PUBYEAR > 1999 AND LIMIT-TO (LANGUAGE, "English").The query led to 836 contributions (the search was first performed on December 4th, 2020 and then updated on February 11th, 2021).Figure 3 describes the steps followed in the selection/exclusion of contributions in the identification phase.

Source Selection
We proceeded with the selection of contributions according to the phases of the PRISMA methodology [51], namely screening, eligibility and inclusion (Figure 3).Concerning the screening, we aimed at discharging contributions (i) out of scope, addressing for example agriculture, construction, geometrical measurement or water; (ii) focusing only on CE; (iii) focusing only on DTs. Figure 3 describes the steps followed for the selection/exclusion of contributions in the screening, eligibility and inclusion phases.To avoid bias, all the phases were conducted autonomously by four reviewers; different results were confronted and discussed, reaching a common agreement.Based on [52,53], we then applied the snowball method, identifying three additional contributions; the contributions were not previously identified as (i) not available on Scopus on the date the last update was performed (n = 1) [54]; (ii) employing specific keywords not included in our query (n = 2) [4,55].The retrieved contributions were classified according to critical dimensions of analysis (Table 2): general information (author; year of publication; journal; document type); bibliometric information (global citations score (GCS) -as by February 11th, 2021; GCS divided by the number of years since publication); content (CE aspects; DTs; type of study; empirical methodology); context (geographical area; sector; size)-see also [56][57][58].Based on the discussion provided in Section 2.1, the CE aspects considered were general, Re-SOLVE, CBM, DIGIT, DISAS, LIFEC, RECYC, REMAN, RESOU, REUSE, SMSER and SCM.Based on the discussion provided in Section 2.2, the DTs considered were general, IOT, BDA, CLOUD, CYB, HVSYS, SIM, AR, ROB and AM.

Data Analysis, Reporting and Using of Results
A critical analysis of the retrieved contributions paves the road to a discussion of the data, identifying key messages and areas for which further additional research is necessary [56].The appraisal was conducted through a descriptive analysis of the results (Section 4) and an evaluation of emerging themes (Section 5).The results deriving from the two analyses are integrated into an overall discussion (Section 6).

Analysis of General Information
The temporal distribution of the contributions shows a more than linear growth, with 85% of them (n = 53) published from 2019 on, highlighting an increasing interest (Figure 4).Considering the journal papers, the most recurrent Journals are Sustainability (Switzerland) (n = 7) and Resources, Conservations and Recycling (n = 6).The distribution shows how the topic has been mainly addressed by sources at the intersection of managementand environment-related areas (Figure 5).In terms of authorship, 188 different authors were identified; 82% of the authors (n = 154) participated in the discourse with 1 contribution, and 16% (n = 30) with 2 contributions.Nine authors contributed with 3 (Bag S.; Nobre G.C.; Okorie O.; Rajput S.; Singh S.P.; Tavares E.; Tiwari A.) or 5 contributions (Charnley F. Moreno M.) (Figure 6).The affiliations were in the United Kingdom (Charnley F

Analysis of Bibliographic Information
As for the impact of the contributions, the highest GCS were 183 [17], 99 [16,98] and 94 [106].The average GCS was about 21, and 89% of contributions received so far less than 50 citations.The GCS was divided by the number of years since publication to better appreciate the breakthrough literature.The highest scores were 146 [17], 33 [98] and 24 citations/year [55].The average score was almost 7, and 79% of contributions received so far less than 10 citations/year.As the first ten contributions according to both the analyses were almost overlapping, the second analysis pinpoints the breakthrough potential of specific conceptual works [55,98] and particularly of the empirical ones by Yadav et al. [113] and Bag et al. [62] (Figure 7).However, numerous contributions from 2020 to 2021 received so far 0 citations, so the list of breakthrough contributions might change in a short time.
Figure 7. First ten contributions according to GCS and GCS/number of years after publication.

Analysis of Content
The most discussed CE aspects were related to CBM, SCM and DIGIT, while among the least discussed, REUSE and SMSER can be identified (Figure 8).Referring to the Re-SOLVE areas, the main addressed ones were exchange and optimize, while among the least discussed, virtualize and share can be identified (Figure 8).As for DTs, the discourse was led by a general perspective on DTs.The most considered single DTs were IoT and DBA; nonetheless, the two of them were strongly interrelated, as IoT can be fully exploited only if the data collected are then processed with BDA [114] (Figure 9).The integrated analysis of the two paradigms, considering how CE aspects and DTs have been integrated, offers sparks for further discussions (Figure 10): SCM was linked to the highest number of DTs and CYB particularly; DISAS and REMAN were linked with SIM and AR; DIGIT has been addressed from a general perspective in terms of DTs.Additionally, the literature has so far considered the impact of more DTs for the optimize and loop areas, compared to the other areas.Details on the specific relationship will be discussed in 5.1.In terms of the type of study, 58% of contributions were theoretical, both review (n = 19) or conceptual papers (n = 20); the remaining share was either empirical (n = 18) or theoretical with a following empirical application (n = 9) (Figure 11).Focusing on the methodology for empirical application (Figure 12), 43% of the contributions (n = 12) employed the case study methodology, followed by surveys or expert opinions.Interestingly, none of the contributions conducted more than 10 case studies, as the majority conducted 1 (n = 6) or 3 (n = 3) case studies.

Analysis of Context
Few contributions considered a specific context.In terms of geographical area, only 28% of contributions considered a specific one (n = 18), with a predominance of European countries (n = 14).As for the sector, 55% of contributions addressed a specific sector (n = 29), and most of them focused on the manufacturing sector in general.As for the size of firms, very few contributions considered a specific one, with 82% of contributions not providing any information.Some contributions nonetheless addressed specifically small and medium enterprises (SMEs) (n = 8), large enterprises (LEs) (n = 3) or both (n = 1).

Digital Technologies Enabling the CE Transition: Emerging Themes
This section discusses the qualitative findings deriving from the literature review, according to the guidelines provided in Table 1 and addressing the role of DTs in enabling the ReSOLVE framework.Although the relationship between DTs and the CE has been largely investigated, and specific connections have emerged, a general complete overview of the relations between the two topics has not been reached yet (Figure 10) [13,17].Nonetheless, interesting points have emerged, starting from a consensus that DTs can act as an enabler for CE.
An in-depth analysis of the insights deriving from the reviewed contributions is offered in Table 3.Each contribution is analyzed according to context and motivation, main contribution, main findings, main limitations and main future research.The analysis provides an overview of the different application areas and outcomes of the extant literature; it is followed by an integrated presentation of the results according to the DTs' potentials in operationalizing the CE transition.
Table 3. In-depth analysis of the content.For the contributions considered for the review analysis, the table reports the following: context and motivation, main contribution, main findings, main limitations and main future research.Limitations in italic refer to limitations related to the specific aim of the present study according to the authors' perspective.

Ref. Authors Context and Motivation Main Contribution Main Findings Main Limitations Main Future Research
[48] Awan et al.
I4.0 and CE pose risks and opportunities to various stakeholders, whose interests and expectations should be understood.
Literature review to identify stakeholders' interests and expectations on how I4.0 can be part of CE transition.
The stakeholders' interests and expectations are a reference point to start a discussion toward I4.0 and CE integration and to shape an organization's strategy for stakeholder orientations.

Systematic protocol limitations (timespan). No focus on specific aspects of the DTs and CE relationship; no focus on operationalization.
Need for empirical research on I4.0-CE relationships.Need to research on CE practices and their sustainability impacts.
Need for better understanding the union between I4.0 and CE.
Investigation of the link between I4.0 and CE, understanding how I4.0 can foster the impact of CE.Thematic and content analysis on grey and scientific literature, to get the perspective of both academia and practitioners.
The current discussion concerns mainly the use of smart services in waste management, resource efficiency and collaboration.There is the need for a better operationalization, also through the conduction of case studies rather than quantitative analysis.
The combination of grey and scientific literature limited the in-depth analysis.More insights from business cases are needed.

No focus on specific DTs-CE relationships.
Future research deriving from the limitations discussed.Need to address and bridge the academic and practitioners' perspectives ('third mission' of universities is encouraged).
The CE transition requires firms to evaluate resource flows, supply chains, business models.The evaluation is critical for high-value manufacturing (HVM).The information on the thermal and mechanical properties and structure of resin available through the system will be a reference point for production engineers.AI allows production engineers to carry out analysis on the data captured by the system.
No practical application.The real-life application could face challenges and require several trialand-error rounds.

Focus only on limited DTs and on a very specific context.
Further research should focus on the conduction of case studies. [72]

Ghoreishi & Happonen
Designing products for circularity is rising in relevance.Parallelly, the adoption of AI in CE solutions increases productivity.
Investigation on how AI can integrate with CE as for the product design phase.
AI helps the optimization of resources for product design, the collection of data on products' lifecycle, the remote monitoring, reuse and remanufacturing of products.

Generalizability of the study limited by sample selection. Focus on limited DTs.
Future research should focus on the identification of barriers and drivers to the adoption of AI for CE, addressing also AI and CE integration in industrial systems as supply chains.
[15] Proposal for an approach to integrate I4.0 in recycling processes.Electric vehicles and their batteries are used as an example.
The information share in supply chains is pivotal for enabling an efficient recycling process.Information can be collected and shared on a marketplace.
Generalizability of the study limited by sample selection.Focus on the recycling process.
Need to enlarge the sample.
[ Future research should provide empirical evidence on the Smart CE, also validating the proposed framework.
As firms are transitioning to CE, technologies allowing the predicting, tracking and monitoring of product's residual value must be identified.
Proposal for an IoT-enabled decision support system for CBMs.Experimental study with a real-world case in the electronic consumer sector.
Products can be tracked and monitored in real-time, through IoT, allowing business analytics.The adoption on the proposed system may support firms in creating more value compared to a linear economy.

Generalizability of the study limited by sample selection.
Focus only on IoT.
Future research is aimed at focusing on the logistics optimization and price and cost prediction.
[79] Moller CE is important within I4.0, and a future ecological and economical model.
Analysis of the digital transformation as an enabler of intelligent manufacturing and its opportunities to CE.
Discussion over the needs for the development of an integrated approach and description of the background for the development.

No investigation of specific DTs nor specific CE aspects.
Need for more interand transdisciplinary research to achieve an intelligent CE.
[ Systematic literature review on robotics in disassembly.
Predefined processes and flexible automation are main research streams.Ample possibilities for integrating the disassembly processes into a superordinate CE information system.

Systematic protocol limitations (search string). Focus only on disassembly and ROB.
Future research will focus on the information processes and system concepts towards an autonomous disassembly system. [84]

Rajput & Singh
The adoption of I4.0 can impact positively on CE and cleaner production.
Proposal for a model for I4.0 set-up to achieve CE and cleaner production, through the optimization of products-machine allocation.
The proposed model optimizes trade-offs between energy consumption and machine processing cost, achieving CE and cleaner production.
The model is developed according to specific hypotheses.

No investigation of specific DTs nor specific CE aspects.
Future research deriving from the limitations discussed.
Companies are urged to re-think their business strategies in view of both the CE and I4.0 paradigms.Presentation of a laboratory application case, testing an electrical and electronic equipment disassembly plant configuration through a set of simulation tools.
Practical demonstration through a laboratory experiment of DTs enabling CE.DTs allow better use of resources, increased production sustainability and benefits along the product lifecycle.
Generalizability of the study limited by the specific context investigated.Focus only on disassembly and ROB, SIM, AM. n.a.
I4.0 and CE are pivotal current topics.They can be described as independent, but overlaps are identified.
Systematic literature review on the relations between I4.0 and CE.A useful double perspective is offered.
I4.0 can generally positively impact the lifecycle management of products and specific insights are dependent on the DTs considered.

Systematic protocol limitations. No focus on operationalization.
Need for empirical evidence on how CE and I4.0 are applied in practice.
[ The analysis of in-service data from automotive components can influence decisions surrounding remanufacture and can lead to significant cost, material and resource savings.

Generalizability of the study limited by the specific context investigated. Focus only on remanufacturing and SIM.
Future research should base on the study to conduct more quantitative and mathematical evaluations.
I4.0 and CE attracted the attention of academia and practitioners, and the connection between them need further investigation.
Application of the situation, actor, process, learning, action, performance linkages framework to analyze the role of I4.0 in realizing CE.
Top managers are essential actors for integrating I4.0 to achieve sustainability, in light of CE.IoT and CYB are pivotal for supporting CE transition.
Limitations related to the possible biased of experts' judgments.
Need for conducting case studies so to understand the roles of digitization and data-driven technologies in achieving the goals of CE. [92] Garcia-Muiña et al.
Eco-design, associated with IoT technologies can help in developing products consistent with CE principles.
Test of eco-design as a tool to define an equilibrium between sustainability and CE in the manufacturing environment of ceramic tile production.Identification of IoT as an enabler for CBMs.
Empirical validation in a manufacturing environment of sustainability paradigms through eco-design tools and DTs, proposing the CBM as an operational tool to promote the competitiveness of enterprises.

Generalizability of the study limited by the specific context investigated. Focus only on IoT.
n.a.[93] Garrido-Hidalgo et al.

Growing need to manage backward materials and information flows in the supply chain, through approaches based on Information and Communication Technologies (ICT).
Proposal for an end-toend solution for Reverse Supply Chain Management based on ICT.Application to an industrial case study regarding WEEE recovery towards CE.
Demonstration of the potential of ICT adoption for Reverse Supply Chain Management.IoT facilitates information management, contributing to CE transition.
Identification of communication bottlenecks that need to be tackled to enhance the reliability of largescale IoT networks.

Generalizability of the study limited by the specific context investigated. Focus only on IoT and CLOUD.
Future research will assess the economic and environmental viability of the proposed approach.
Item-level identification can foster disruptive innovation, enabling CBMs.
Proposal of a method to facilitate IoT for building a product passport and support data exchange, enabling CE.
SmartTags can be used in CE for unique item-level identification and detection of environmental parameters.
The solution is evaluated according to specific hypotheses.
Focus only on IoT.
Need for further research to test all the hypotheses.
Potentials to combine I4.0 and CE to enhance the sustainability of manufacturing sectors.
Exploration of the I4.0 factors accelerating the sharing economy.Investigation through a case of electric scooters in Taiwan.
I4.0 is an enabler for sharing economy.I4.0 technologies are helpful to overcome specific barriers to CE adoption.

Generalizability of the study limited by the specific context investigated. Focus only on IoT and CLOUD.
Need to approach CE with a holistic, policyoriented approach. [100]

Rajput & Singh
An integrated I4.0-CE approach can increase efficiency and optimize the entire value chain.Thanks to I4.0, possible technological barriers to the CE transition might be overcome.

Identification of I4.0 barriers to CE. Prioritization of barriers and identification of contextual relationships among them through Interpretive Structural
Modelling.
The main barriers are process digitalization, sensor technology and design challenges.An I4.0-CE approach would allow operations management sustainability, optimizing production and consumption, while also providing opportunities for customization.
Possible bias and subjectivity in the identification of contextual relationships.

No investigation of specific DTs.
Future research should provide more detailed and empirical evidence on barriers.
[ Identification of the linkages between the phases of a product lifecycle and the design levels of business engineering.

No investigation of specific DTs.
Future research should better investigate the different success factors.Identification of systems and methods used in waste management sector and of technologies applied in other sectors that could be relevant as well.
Robotic-based sorting and lifting systems in waste management are pivotal, as they also partially replace humans.Limitations can be identified, material-and technology-wise.

Focus only on recycling and IoT and ROB.
Future research should address the sensors needed for a successful application of I4.0 for waste management.[ Proposal for a system allowing the planning and organization of processes, so to minimize raw materials' consumption and reduce negative environmental impacts.

Focus only on SCM.
Future research should focus on simulation models for the adoption of the proposed system.Future research should focus on barriers and drivers to the CE transi-zation for CE transition need investigation.
efficiency in the German industry.
dressed primarily to improve efficiency in the manufacturing process.
No investigation of specific DTs.
tion, while also evaluating the economic benefit from the adoption of DTs and CE.
Opportunities to apply the CE to the rapidly changing paradigm of I4.0 need investigation.
Systematic review of the empirical literature related to DTs, I4.0, and circular approaches.
Proposal Future research should provide more detailed and empirical evidence. [ Redistributed manufacturing and CE can potentially disrupt current models of consumer goods production and consumption.
Exploration of digital intelligence and redistributed manufacturing as enablers of CE.Analysis of literature case studies.
The integration of DTs can enable the distribution of knowledge, customization and CBMs.Circular innovations support more regenerative and resilient systems of production and consumption.

Findings based only on secondary data. No investigation of specific DTs nor specific CE aspects.
Need for empirical research to further validate the findings.
[112] Reuter Process metallurgy support CE; the digitalizing of the material production could provide additional support.
Evaluation of the different possibilities and application for the metallurgical IoT.
Identification of opportunities, limits, tools, and methods of process metallurgy and recycling within the CE, through the adoption of DTs.
Generalizability of the study limited by the specific context investigated.

Focus only on IoT and
BDA.
Future research should focus, among the others, on the role of the disruptive CBMs.

Digital Technologies Enabling the ReSOLVE Framework
A narrow group of contributions investigated the relationship between the DTs and the overall ReSOLVE framework (Figure 10).The contributions were mainly review or conceptual papers.They considered one DT or a limited set of them, paving the path for more integrated analyses.From this perspective, Nobre and Tavares [80] linked some aspects related to IoT and BDA to the different areas of action of the framework.BDA and their requirements for appropriate applications in the different areas were also discussed [55].IoT was also considered together with CLOUD and AM [15,17], and a framework for fostering their adoption was also proposed.Lastly, different examples of CYB applications in the ReSOLVE framework, focusing particularly on the benefits related to traceability and security, have been provided [75]; however, the proposed applications were still at a pilot or planning stage.
The reviewed contributions are nonetheless mostly focused on specific CE aspects associated with the ReSOLVE action areas (Table 1).

DTs enabling the Regenerate area
The Regenerate area has been so far connected to a limited series of DTs (Figure 10).Regenerate area could benefit from DTs thanks to the application of IoT in the form of sensors for the collection of data and BDA for the elaboration of the collected data [17].A decisive positive impact of IoT on the product lifetime extension is underlined in terms of monitoring, control and optimization, allowing additional support and value to the customer [95].As a part of IoT, the use of Smart Tags for building a product passport and enabling data sharing and exchanging is supported [94].The use of BDA is then necessary for a proper elaboration and use of such data, facilitating the decision-making process [89].Riesener et al. [102] detailed different phases of the lifecycle, namely manufacturing, usage and reutilization/recycling.As for the manufacturing phase, CBY, IoT and HVSYS can help to solve information asymmetry; concerning the usage phase, CYB and particularly blockchain might support the handover to different customers, also enabling the traceability of the product and the acquisition and verification of related data [115].Regarding the reutilization/recycling phase, BDA might allow different cycles, fostering a reverse logistics system and waste management [116].

DTs enabling the Share area
The research over the use of DTs in support of the Share area appears rather limited (Figure 10).IoT allows the monitoring and tracking of the use and condition of products, thus enabling reuse [95].As a large amount of data would be collected, BDA and CYB become again of fundamental importance to manage the complexity [102].The collection of data on the product condition and the related decisions for reuse would allow better cooperation among the tiers of the value chain [66,89].

DTs enabling the Optimize area
The enabling role of different DTs in terms of Optimize area received rather good attention (Figure 10).Nonetheless, despite evidence that DTs can support resource efficiency [12,66], firms still lag as for DTs adoption and exploitation [109].A pivotal role is played by IoT for monitoring, control and optimization [95], allowing also the identification of resource waste in real-time [92].The IoT would then require the support of BDA [102,112].Nonetheless, the collection and analysis of data could be insufficient, and the use of CYB is suggested to share the product information among the different stakeholders, while also facilitating the paperwork activities and the checking of the status of the products along the supply chain [69].
From a general perspective, the role of DTs is also studied concerning supply chain management [101].As the management of inbound and outbound logistics is particularly relevant, firms might benefit from DTs applications in procurement and logistics, also helping build the capabilities needed for collecting, processing and sharing information [62,63].From a practical viewpoint, IoT can allow the real-time evaluation of the product value along the tiers [78], with an exchange of the data stored in a CLOUD inventory [88,93].
DTs could also support reverse logistics, with a particular relevance of BDA, on data collected with an HVSYS perspective [108], possibly fostered by AM production system [68].The transparency and security of data exchange and any type of digital transactions can be guaranteed by the use of CYB; from a larger perspective, the use of CYB could also support supply chains in making their practices more transparent, secure and correct [54,61].Possible different configuration scenarios can be then analyzed thanks to SIM [97].As a last remark, preliminary insights on the combined support from DTs and CE to enhance sustainability started being discussed [61,88].

DTs enabling the Loop area
The Loop area can benefit from the adoption of different DTs (Figure 10).Particularly, a good variety of DTs proves to foster actions related to the disassembly of products.An interesting role is played by AR, which could be useful in planning the disassembly sequence, as it would allow the visualization of all the information and equipment needed in the process, besides helping in training operators [96].ROBs also have an interesting role in the disassembly process [85], although issues in terms of economic feasibility may pose limitations [83].As for the determining and optimizing of the disassembly process, both SIM [85,91] and BDA for mining a repository of disassembly processes [70] are considered as possible options.
The remanufacturing process requires different data related to the product, i.e., its status, maintenance history, disassembly and reassembly [117].From this perspective, the use of IoT via sensors would be helpful to track the product history, through real-time monitoring that could bring positive effects in different processes [89,118].However, although few cases can be spotted where IoT is relevant for looping strategies, empirical studies show that IoT is not largely used for product remanufacturing [95].Additional opportunities have been conceptualized from the integrated use of IoT and AM, but the realization has not been demonstrated [68].AM nonetheless can contribute to sustainability, presenting lower cost related for example to set-up, and can play an important role in the loop area when the workload is distributed along the different tiers of the supply chain [68,119].For the latter point, the use of an HVSYS would allow real-time data management throughout the entire chain, thus facilitating the loop strategies [68].Lastly, SIM can assist remanufacturing processes, as discrete event simulation models are essential to determine the quality of a product [91].
As for the recycling process, IoT would provide benefits in terms of monitoring and tracking [103], enabling looping [71,95].Two relevant aspects emerged as connected to the adoption of IoT: first, to make proper decisions, data should be collected with an HVSYS perspective [74]; second, as many data would be collected, BDA becomes fundamental to manage the complexity [102].Lastly, the use of ROB would facilitate the recycling process, while also bringing benefits from a social sustainability perspective [103].Additional insights are provided highlighting a strong correlation between the adoption of recycling practices and the adoption of BDA, ROB and AM, suggesting also that I4.0 might allow greater integration among the partners of the value chain [66].Nonetheless, several issues emerge trying to link the concept of CE with the one related to industrial systems; incorporating CE into industrial networks requires a change in the economic paradigm [120] that would be allowed only by a strong willingness of all the involved partners to embark in this transition [121].The main obstacle and critical resource for the transition is the trust among the partners [122], which becomes even more pivotal if the transition is enabled by DTs [123].Additionally, moving from a single firm to the value chain and then the industrial system would require the identification of the best set of DTs to use [124].

DTs enabling the Virtualize area
The enabling role of DTs towards the Virtualize areas has been so far investigated from a rather limited perspective (Figure 10).As for SMSER, the discourse has been approached from a general viewpoint, without much detail on specific DTs.Exact points have arisen in terms of the need of being able to collect data to monitor and evaluate the conditions of the product, using for example IoT [89], complemented with CLOUD [99] and DBA for the analysis [105].Additionally, also AM can be interesting for the customization of products based on interactions among different tiers of the value chain [68].The action area can be supported by IoT, as they could foster the relationship and communication between organizations, suppliers and customers [68].

DTs enabling the Exchange area
The area has been largely addressed mainly from a general, broad and theoretical perspective (Figure 10), understanding how the presence and adoption of new technologies could foster and support the transition towards new update CE practices.DTs could indeed offer a solution for core data records concerning a sustainable product and material database [79].Opportunities can be found at different levels, as the optimization of the resources use, the engagement in business models enabled by software development, the share of information on a network level, the creation of infrastructures supporting the tracking and monitoring [104], while also fostering cleaner production [84].Although so far, the discourse has been mainly theoretical, some first empirical applications can be found for specific CE aspects, as redistributed manufacturing [111].
Concerning specific DTs, the discourse mainly developed around the adoption of IoT and BDA [16,67,77,112], with the latter playing a fundamental role also in terms of predictive analytics [64] with the support of CLOUD [67].Positive impacts were also observed concerning AM, as it could easily support CE strategies focused on materials [76].
At this stage, contributions also focused on the identification of barriers and challenges to and for DIGIT [67,100,104].Some of the barriers refer to organizational aspects (lack of competencies, need for coordination, need for technical development), to economic aspects (financial and operational risk), as well as the digitalization process itself.Particularly, the adoption of DTs can be strongly hindered by organizational resistance deriving from both the employees and the management, who can oppose the change within their organization [125] and might lack specific competence and skills [126]; particularly, the role of managers has been considered of fundamental importance to support the integration of DTs in the light of CE [12].Additional challenges seem to emerge concerning the context of the investigation, with developed countries facing mainly issues related to the low maturity level of the desired technology [127] after national strategies and policy have been formulated [128], and the developing countries still struggling with the setting of proper standards and legislation [127].Specific barriers in the context of emerging economies were indeed pointed out in terms of the macro environment [90].Additionally, an important role is played by the availability of the technologies, also from an economic perspective [17].In this scenario, it is pivotal to identify the best drivers to overcome the barriers [129].

Digital Technologies Enabling the CE Transition: Further Insights
According to the overview provided in 2.1, the ReSOLVE framework entails major circular business opportunities [27].The literature has largely addressed the role of DTs as possible enablers of CBMs; however, the important operational role of the ReSOLVE framework has not been considered.For having an overview as complete as possible, we considered also this general viewpoint.From a broad perspective, DTs allow the industry to embrace innovative, productive and sustainable CBMs [86,105].A central role is played by the exploitation of data [13] and consequently data collection, integration and analysis, using IoT, CLOUD and BDA [60].DTs would also allow greater involvement of customers in the definition of CBMs [9], leading for example to customized services [97].
Analyzing specific DTs, IoT is undoubtedly among the pivotal ones: on the one hand, it can support the definition of servitized CBMs [73]; on the other hand, it can advance the tracking, monitoring and control of products [73,95], favoring a real-time analysis of the product's residual value [78].The adoption of IoT implies the need for a good quality of data and appropriate data management [73], so that BDA becomes fundamental [15,107].Additionally, IoT and BDA together can support specific aspects at each product life cycle stage (as product design; marketing activities; monitoring and tracking of the product; technical support and maintenance; product optimized use, upgrade and renovation) [106].To effectively adopt IoT and BDA, some aspects are necessary [72] such as (i) the collaboration with stakeholders and particularly customers to obtain the data, (ii) the capability of workers to analyze and manage the data and (iii) the consideration of impacts from a sustainability perspective, thus including economic and social aspects-given the high cost of DTs, and the strong relationship between product and customer satisfaction [130,131].Lastly, within this framework, CLOUD is fundamental for the storage and share of data [87].The literature also offers insights into the relationship between CBMs and AM as an enabler of CE.AM emerges as capable to increase productivity and manufacturing freedom on demand, targeting the needs of each customer, while also enhancing sustainability, with economic, environmental and social implications [98].

Digital Technologies Enabling the CE Transition: Discussion and Open Issues
The analysis of the literature confirmed the relevant role of DTs in enabling and supporting the CE transition.The trend in terms of year of publication (Figure 4) underlines how the research on the topic is relatively young, as also noted by previous research [21,82,104].The geographical distribution of the authors is showing a global interest in the topic by both developing and developed countries (Figure 6), as also previously underlined [4].The descriptive analysis of the different CE aspects (Figure 8) highlights how the research is still mainly focused on specific aspects of CE, and it is not integrated into a more structured and operative framework, as the ReSOLVE one [16].In this way, DTs are related to specific CE aspects or processes, and it is difficult to have a complete overview of all the benefits that DTs could bring to the overall CE transition (Figure 10).As also emerged from the descriptive analysis, the discourse is still mainly driven by theoretical contributions, and particularly literature reviews, so that advancement from both a conceptual and (mostly) an empirical perspective is strongly recommended (Figure 11).The urgency is also underlined by the breakthrough potential of the empirical research conducted on the topic (Figure 7).Particularly, considering the insights that emerged from the present review (see also Table 3), the following issues are worthy of note and urge for additional research efforts.
Integrated and holistic perspective on the DT-CE relationship.Shortcomings can be identified in the evaluation of the relationships between DTs and CE from both sides.Regarding DTs, the largest share of contributions focuses on one or a limited set of DTs (Figure 9), while the contributions addressing DTs in general terms mainly provide few examples on specific DTs or applications.Nonetheless, DTs are for their nature interconnected, and it may not be possible to adopt a DT without at least a partial presence of another one [104].The research on the integration of the different DTs shows indeed a growth potential for better investigating their role in the CE transition [48,110].As for CE, the literature is still mainly focused on specific CE aspects, not considering a more integrated approach (Figure 8).Although the literature largely recognized the enabling role of DTs, there is an urgency to investigate how they enable the transition from a more operative perspective [107], as the one offered by the ReSOLVE framework [16].Additionally, as emerged from the review, specific ReSOLVE action areas and aspects of CE are more investigated than others (Figure 8), leaving ample room for additional research.To move towards enhanced CE, firms should adopt CE practices.As DTs enable the CE transitions, it comes directly that DTs could also enable and support the adoption of specific CE practices.The investigation of the relationship should consider the intensity with which a specific DT could impact the CE transition, not only from an overall perspective but also regarding specific action areas and specific practices that a firm could implement-see, for example, [132].Such analysis would make it easier for the industry to understand if, how and to what extent the adoption of specific DTs could impact the CE transition, possibly allowing them to better organize their resources and concentrate their efforts towards the adoption of those DTs that could be more efficacious.
The decision-making process.The CE practices would have to undergo an adoption process that could be influenced by several factors, as demonstrated for example for industrial sustainability [58,130,133].The evaluation of these factors would be of fundamental importance to better understand how and to what extent DTs can enhance specific CE practices.In particular, a holistic investigation on the following points is advised, understanding their role in the different phases of the adoption process of CE practices, and how they could change according to the action of different DTs:  Barriers to CE transition: identification and evaluation of the inhibitors of the adoption process [9,17];  Drivers for CE transition: identification and evaluation of the fostering factors for the adoption process [17,66];  Performance measurement: identification and evaluation of the performance reached after the adoption; fundamental for this aspect would be the identification of how the performance could be gauged [17,59].Another important aspect to consider is the evaluation of performance beyond the ones strictly related to CE.As introduced in the previous section, some authors started investigating the performance related to the overall sustainability derived from the adoption of CE practices supported by the DTs, see for example [54,98,134].However, despite the common agreement, additional research seems to be necessary to better determine the relationship between DTs and industrial sustainability [135,136];  Contextual factors: identification of those contextual factors, as geographical area, sector or firm's size that could influence the adoption process [137] and that so far appear still limitedly investigated (see Section 4.4.);previous research showed a pivotal role of the firm's size, particularly when SMEs and LEs are confronted [138][139][140];  Digital maturity level: evaluation of the impact of the digital maturity of the firm on the outcomes, as it might represent a quite important influence [98,141];  CE management: evaluation of the impact of how CE is managed within the firm, as it might influence the outcomes [137].For example, the presence of an environmental management system demonstrated to strongly support the CE transition [142]; as a clear predominance for a heterarchical control for DTs has been underlined [143,144], the debate on whether a centralized or decentralized system would be better for environmental-related aspects is still open [145,146].
Empirical research.As abovementioned and shown in the descriptive results (Figures 11 and 12) and highlighted by the in-depth analysis of the content (Table 3), the largest share of the published contributions employs a theoretical approach.This urges for more empirical research, which is also highlighted in previous literature [15,77,87,110].Although an increase in empirical studies can be appreciated in the latest years (Table 2) [82], there is still ample room for providing practical demonstrations of the impact of DTs on the CE transition.To deepen the understanding of the relationships, the adoption of the case study methodology is suggested, providing more qualitative than quantitative evidence, but allowing a deeper analysis of the context under investigation [101,147] (see also Table 3).As some contributions employed the case study methodology, the investigations present less than five case studies, with the largest share of contributions focusing on one case study (Table 2; Figure 12).To confront an already existing theory toward an empirical application and structuring the theory in light of the observed results, a larger number of case studies is therefore suggested [148,149].
The role of industrial systems and stakeholders.Moving from a micro to a meso-level of analysis, DTs have been proven to facilitate the cooperation and connection of firms, foster industrial symbiosis and help build a collaborative environment to promote the CE [104].Regardless of the provided insights on the possible influence of DTs on SCM, it is suggested to conduct specific studies investigating industrial systems from the perspective of all the involved firms, not only a focal/single one [131,150] (see also Table 3).

Conclusions
The present study critically reviewed the literature on the role of DTs in operationalizing the CE transition, shaping the analysis according to the ReSOLVE framework.
Our analysis indicates a broad focus on the topic, yet there is still the need to tackle it in a more integrated and holistic manner.The discourse is mainly focused on single DTs enabling specific CE aspects; thus, it is tough to have a complete view of all the possible DT implications on the overall CE transition and operatively address the transition itself.From this perspective, the paper suggests interesting directions for further research, aimed at addressing the operationalization of CE through DTs, with an integrated and holistic perspective.
The present study offers contributions from both theoretical and managerial viewpoints.First, we analyzed 66 literature contributions using a comprehensive list of axes for the evaluation: these axes could be useful for scholars and managers alike as a reference guide to continue the exploration of the topic.Second, we provided an analysis of the previous literature according to the axes of evaluation, spurring interest in future research.Third, we suggested the need for additional research on the topic; such research should provide a more integrated and holistic view on the topic itself, supported by strong empirical evidence.Leveraging on this, further research from academia is fostered, so to support practitioners in understanding the best manner to exploit the enabling potential of DTs.
We conducted our analysis following the principles of ethic, quality and accuracy.Nonetheless, some limitations should be highlighted.First, we conducted our study considering only Scopus as a scientific research database, and different findings may be obtained using other databases.Second, as the role of DTs as an enabler of CE is a current hot topic in the managerial and academic debate, the number of studies on the argument is constantly increasing, and the specific time frame we used could have excluded some relevant recent contributions.Future research should be thus directed to consider the abovementioned limitations, while also investigating the evolution of the research topic.

Figure 3 .
Figure 3. Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) methodology.The figure reports the different phases of the PRISMA methodology.

Figure 4 .
Figure 4. Distribution of the reviewed contributions according to publication year and type of document.

Figure 5 .
Figure 5. Distribution of the reviewed contributions according to the most frequent journals.

Figure 6 .
Figure 6.Distribution of the most prolific authors according to the country of affiliation.

Figure 8 .
Figure 8. Distribution of the reviewed contributions according to CE aspects considered.

Figure 9 .
Figure 9. Distribution of the reviewed contributions according to DTs considered.

Figure 10 .
Figure 10.Heatmap of the distribution of the reviewed contributions according to CE aspects and DTs considered.

Figure 11 .
Figure 11.Distribution of the reviewed contributions according to the type of study.

Figure 12 .
Figure 12.Distribution of the reviewed contributions according to the methodology employed for the empirical application.

[ 103 ]
Sarc et al.I4.0 are implemented in the field of waste management to achieve CE.

[ 109 ]
Neligan Opportunities and challenges of digitali-Empirical findings on the importance of digitalization to improve material Opportunities deriving from DTs are limitedly exploited and ad-Generalizability of the study limited by the specific context investigated.

Table 1 .
Linkage of CE aspects with ReSOLVE's actions.

Table 2 .
Source evaluation.For the contributions considered for the review analysis, the table reports the following: general information: author, year of publication, journal, document type (JP: journal paper; CP: conference paper; BC: book chapter); bibliometric information: GCS, GCS/number of years since publication; content: CE aspects; DTs; type of study (R: review; C: conceptual; E: empirical); empirical methodology; context: geographical area, sector, size.