Introduction

The world population in 1950 was 2.5 billion people. As of August 2016, it was estimated as 7.4 billion. By 2050 it will be nearly 9.8 billion, an increase of 392% in 100 years (United Nations 2015). Natural resources consumption is also significantly growing, as a consequence of its current observed positive correlation with country’s per capta GDP (World Bank 2014; SERI and Dittrich 2014) which indicates that the country’s wealth has a direct influence on its natural resources consumption. Proportionally but in the opposite direction, we find natural resources availability to support this economic growth. For example, assuming no changes occurring in global economic growth tendencies and no actions taken, oil reserves will end by 2057 and natural gas, by 2070 (Guardian 2011). Moreover, resources consumption results in waste generation that returns to the environment. Despite known market and government regulatory actions in order to promote recycling initiatives, only a very low part of recyclable wastes are actually recycled (Ellen MacArthur Foundation 2016). Currently there are nearly 150 million tons of plastic on the oceans and by 2050 there will be more plastic than fish (World Economic Forum 2016). One of the main causes of this situation is the current linear economic model (“take, make, use, dispose/waste”), present since the industrial revolution and considered nowadays an unsustainable practice. This calls for an alternative model and circular economy (CE) is being advocated as a solution (Andrews 2015).

The CE model consists in a continuous and positive development model focused on preserving and enhancing natural capital, optimizing resource yields and minimizing system risks, by managing finite stocks and renewable flows, and keeping products, components and materials at their highest utility and value at all times (Ellen MacArthur Foundation 2016). Despite the concept existence for decades and the benefits of the model being clear, the acceptance of CE business models is still modest, as the key lever for change relies not only on rational but also on non-rational motives of consumer behavior, which includes habits and routines of individuals (Planing 2014). The emergence of new technologies leveraged by the Internet, mobility, tagging/tracking, low costs etc., along with changes on consumer behavior (e.g. access over ownership) might have been what was missing. Such technologies can be summarized into two complementary subjects: big data and internet of things (IoT).

Data has never been so important for enterprises/other entities as it is now and may become the biggest trading commodity in the future (Xiao et al. 2014). IoT, leveraged by big data, is considered by specialists as the next Internet evolution, allowing a huge step towards data gathering, analysis and distribution, which can be converted into useful information (Evans 2011). There are ongoing projects such as reducing the gap between the poor and the rich and better resource distribution, and helping people to be more proactive against nature events (earthquakes, tsunamis, polar ice cap melting etc.) for example, all based on IoT/big data (Hart and Martinez 2015; Evans 2011). It is also noticed that the plateau of productivity for these technologies will occur in the next 5–10 years (Gartner 2015), fostering researches on the subject.

By understanding the basics behind the concepts of CE, big data and IoT, it is not hard to realize how integrated they can be with each other, and their role on the development of a better society. Selected industry cases presented on the document show the potential of this integration and how it will change the world from the environment preservation and human behavior perspectives.

The purpose of this research is to understand, from a bibliometric perspective on scientific literature, the state-of-art of CE and the role of IoT/big data in this context, and compare it with some of the current initiatives being undertaken by the corporate world, also allowing the gap between industry/private sector initiatives and scientific research to be better understood.

As this is a preliminary but also unprecedented study, we expect the results/conclusions of this research will motivate future exploratory/practical studies in the field of CE leveraged by cutting-edge technologies, and with a higher level of integration between industry/private sector and scientific research, fostering and accelerating researches in the area with the use of methodologies that promote practical actions and creating knowledge/theories at the same time, such as action researches (Coughlan and Coghlan 2001).

Main findings indicate there are only a few CE studies (<0.25%) considering the application of IoT/big data, although a growing interest has been observed on the last three years. The results further indicate a research agenda linking IoT/big data to CE, suggesting areas currently in development, such as carbon emissions reduction, smart cities, water economy, waste management and renewable sources. Cloud computing has been an additional enabler. There are also areas to be explored, such as bioeconomy and carbon biofixation. Also, a disconnection between industry/private sector and scientific research seems to exist, as the scientific studies found are not related whatsoever to the important industry/private sector example studies used in this research.

The reminder of the paper is organized into five sections, starting with insights on academic researches × industry/private sector initiatives, followed by a methodology section, which includes the data collection steps used on the bibliometric research. The last three sections contain the results, policy implications and conclusions of the study.

CE with big data and IoT in the industry/private sector and the academy

The CE principles adoption, which in most cases will be achieved with the support of technologies such as big data and IoT, can create net benefits for the economy of about €1.8 trillion by 2030 only in Europe, representing €0.9 trillion more than the current development path—the traditional “take, make, use, dispose/waste” linear model, and achieve savings of up to USD 1 trillion on materials savings (Ellen MacArthur Foundation 2015a). Nevertheless, the practical effects and results measurement of big data and IoT remain a question mark, as they are in their early days, so the potential for value creation is still unclaimed (Groves et al. 2013), which may justify somehow the low amount of academic work that combines both subjects.

Some recent scientific publications have been questioning the low amount of researches on the field of environmental sustainability with the support of information systems. The Journal of the Association for Information Systems, for example, made a call for papers on IS solutions for environmental sustainability and managed to publish only one due to the low amount of papers submitted and researches scope (Gholami et al. 2016). Although the demand for scientific research in the area is high and continues to grow, scholars seem to lack what is needed to fill the gap between the researches being undertaken and practical actions to solve environmental problems. Some barriers for researchers seem to exist, as described in a recent study (Gholami et al. 2016): incentives misalignment, the low status of practice science, data analysis poverty, identification of the proper research scope, and research methods. Given these barriers on conducting this kind of research, one possible path to be followed by scholars is the application of action research as the preferred method for such turbulent research environments (Daniel and Wilson 2004), in order to approach practice (which seems to be far more advanced on CE with big data/IoT) with theory. In addition, along with the action research method, the application of concepts of engaged scholarship, which overcomes the barriers of individual researches and also allows the partnering and gathering inputs/valuable advice from practitioners, customers, users, sponsors and other stakeholders, can generate outcomes that exceeds the traditional knowledge production way (Van de Ven 2007). This may allow social researches to result into valuable insights not only for the academy but also to practical use for the industry/corporate world.

In order to set some comparison parameters to understand the distance between current academic publications for CE with big data/IoT and the corporate world, we selected some examples of cutting-edge industry/private sector initiatives, which also illustrates the importance of those concepts when bound together, for different purposes/applications.

Philips lightingFootnote 1

The Spanish Volkswagen subsidiary (SEAT) established a partnership with Philips Lightning to replace its exterior light sources by connected LED lighting, which is more efficient and sustainable, enabling the company to save 80% on energy consumption, amounting to over 900 MWh per year. And the relationship model is also new and different. They implemented the light as a service model, so all the installations are owned by Philips, who provides the lighting services to SEAT. This represents a CE initiative that promotes the “access over ownership” model (Ellen MacArthur Foundation 2013). Upon full project conclusion, more than 1200 fixtures will be replaced by a more environmentally efficient system. With the use of technology, lightning needs can be controlled and monitored online and real time, including inventory management support with the geolocation system of all fixtures (IoT). This enables the partnership to fully monitor and control the installations, obtain direct information from each light fixture (big data) and supervise improvements in energy expenditure management as well as in maintenance, generating savings for both partnership and environment.

Cisco’s sport shoesFootnote 2

A collaboration of Cisco, Cranfield University and The Clearing institutions resulted in a project to explore in detail a specific real-world application combining big data, IoT and digital technology to drive a new, circular model for production and consumption of sport shoes, which allowed the implementation of a re-distributed manufacturing (RdM) system in UK. This term is used to describe the transformational shift from centralized to decentralized (usually smaller-scale) manufacturing, powered by digital technologies. Consumers deliver the orders by scanning their feet with a cell phone app and use virtual reality to assembly their shoes based on hundreds of options according to their requirements (visual, comfort, performance etc.). 3D printers are used to assembly the modular and customized product in a location close to the customer, so the manufacturing occurs decentralized, allowing lower distribution costs and raw materials savings. Also, consumers don’t actually own the products, as they pay an ongoing subscription to the vendor. Products come with an intelligent component (IoT) that tracks location/shoes condition (according to consumer accordance) and identifies replacement/upgrades needs. Upon life-cycle end, consumers return the shoes for remanufacture in a take-back model to a location close to their home/work. The shoes’ modular design allows them to be easily disassembled for refurbishing/recycling. Bottom-line: low production costs, highly customized products, circular economy application, “access over ownership” model, satisfied consumers, big data and IoT acting as enablers.

Arup’s circular buildingFootnote 3

The company used big data/IoT technologies to maximize utilization of components and materials on a building construction. The facility was designed for easy disassembly, and with the use of non-toxic and pre-fabricated components made to be taken apart without damaging them. They used an innovative digital tool called building information modellingFootnote 4 (BIM), which allows data communication among stakeholders during all phases of an asset’s lifecycle. It envisages the virtual construction of a facility prior to its actual physical construction, in order to reduce uncertainty, improve safety, work out problems, and simulate and analyze potential impacts. Contractors from every trade can input critical information into the model before beginning construction, with opportunities to pre-fabricate or pre-assemble some systems off-site. Waste can be minimized on-site and products delivered on a just-in-time basis rather than being stock-piled on-site. BIM also enables optimized design processes and supports the efficient performance and maintenance of buildings. By incorporating information on materials, it can help communicate any negative externalities as well as opportunities for recycling and remanufacture. Moreover, as BIM is also a database containing all components used on the construction, the building also acts as a “material bank” for components reuse.

Other companies that have been promoting this revolution include: Uber (currently the world’s largest taxi company that owns no cars, but fosters the higher use of the current car fleet rather than putting more cars on the streets), AirBnB (largest accommodation provider without one single room on its own, but helping people better use their available spaces), all with the support of big data and IoT.

Those concepts of CE, big data and IoT, when bound together, offer a large set of opportunities for both business and society. The use of enhanced tracking/tagging capabilities, for example, presents significant economic opportunities to save materials, reduce waste and make use of resources previously considered to be disposed. Therefore, IoT plays a key role in providing valuable big data about things like energy use, under-utilized assets, and material flows to help make businesses more efficient (Ellen MacArthur Foundation 2016).

General overviews on basic concepts of CE, big data and IoT are presented in Appendix 5.

Methodology and data collection for the bibliometrics analysis

The first step consisted in identifying publications from a robust and reliable database. Scopus was the chosen database as it is considered one of the largest abstract and citation databases of peer-reviewed literature, including scientific journals, books and conference proceedings, all part of the scope of the study. The research scope focused on publications in a ten-year time interval (2006–2015).

The research initially retrieved publications regarding only CE to serve as the baseline for the comparisons. The initial query on the term “circular econom*” in the titles, abstracts and keywords resulted in 915 documents. Nevertheless, publications containing terms and expressions semantically different but with the same meaning/or a subset of CE were not retrieved by the query. Those terms include “Circulatory Economy”, “Circular Supply Chain”, “Circular Ecology” etc., and they had to be added to the research filters. Additionally, publications fostering CE initiatives and opportunities do not necessarily utilize this expression within its body, partially because the regular use of the term is relatively recent (18 occurrences in 2006 × 151 in 2015, an increase of >800% according to the Scopus database).

Many CE-type studies make use of other key expressions along with terms like “sustainability” or “sustained development”. For example, a study focused on carbon dioxide conversion into organic compounds that can be synthesized into hydrocarbon fuels (Whipple and Kenis 2010), cited more than 230 times, is indeed a CE study but the expression is not used throughout the document. A keyword for this study would be “CO2 Reduction”. On the other hand, some CE terms may be applied to other concepts not related do the research scope. One example is the one of the CE’s schools of thought term “blue economy” (Pauli 2010; Ellen MacArthur Foundation 2015b), also used in studies regarding economic perspectives and governance between ocean and land (Smith-Godfrey 2016) and confirmed as a separate subject in discussions/events (e.g. Rio + 20) (Silver et al. 2015) and for that reason considered as out of scope for this study.

The sensibility of the CE terms required a more detailed approach for the keywords definition. Therefore, a taxonomy research was conducted in order to map the keywords for the research and a number of publications were found with different definitions and classifications, which led to the need of a broader keywords search. For example, although the classical 3R concepts (reduce, reuse, recycle), closely related to CE, are still up to date (Terazono et al. 2006; Agamuthu and Fauziah 2011; Tian and Chen 2014; Nagalingam et al. 2013), some authors have included a 4th R standing for recover or reclaim or refuse or reorganize or refurbish (Nagalingam et al. 2013; Rehman and Shrivastava 2014; Hickey et al. 2014). 6R definitions were also found, adding new terms (Kuik et al. 2011; Yan and Feng 2014). Furthermore, despite the existence of a taxonomy for the end of life (EoL) projects and processes based on the strategies starting with the prefix “Re” (e.g. reduce, reuse, repair, recycle etc.) (Sihvonen and Ritola 2015), it does not intend to cover comprehensively all expressions related to CE. Therefore, in the absence of a reliable and up to date taxonomy for CE, the complete collection of keyword terms was obtained as follows: from the original “circular econom*” search results, all data from the keyword field was extracted, generating a list of 2488 additional unique terms. The list was then reviewed according to the steps shown on Fig. 1, resulting in 116 terms. New terms were added based on literature review (cradle-to-cradle, biomimicry, regenerative design, resource recirculation, regenerative econom*, restorative econom*), generating a final version of 122 items.

Fig. 1
figure 1

CE keyword search mapping process

Full size image

The query logic was then formulated, according to Fig. 2. A special treatment was given to the keywords “Reduce”, “Reuse”, “Recycle”: taken together, they represent the 3R concept, but when searched separately, some out of scope results may be retrieved. The final CE query applied on Scopus Database containing all research terms is available in Appendix 1.

Fig. 2
figure 2

CE query logic

Full size image

Previous studies indicate the need to include proceeding papers in citation impact comparative researches rather than relying only on journal articles because of the significant influence this document type has on citation impact (Ingwersen et al. 2014). Therefore, considering the inclusion of scientific journals, books and conference proceedings, the final CE query resulted in 30,557 documents.

Similar treatment was given to the search criteria for “Big Data” and “Internet of Things”. In contrast to the previous search, in this case more consistent taxonomies were identified (Debortoli et al. 2014; Bajaber et al. 2016). However, some terms when searched separately, such as “Machine Learning” presented results out of the scope and had to be combined with other terms. The resulting query logic is represented in Fig. 3.

Fig. 3
figure 3

Big data and IoT query logic

Full size image

The final big data and IoT query applied on Scopus Database containing all research terms returned 32,550 documents and is available in Appendix 2.

Two main bibliometric indicators were used in this research: CPP (citations per paper) and PNC (percentage of non cited papers) (Van Raan 2005), along with absolute numbers for publications and citations, grouped by country.Footnote 5 Standard deviation and maximum number of citations were also used as auxiliary fields to better results description. When assessing publication impact with the use of citations per paper (CPP) and percentage of non cited papers (PNC) indicators, we should consider that the citation peak of an article occurs around the 4th full year since its publication (Li and Ho 2008) and the 6th year (Hassan et al. 2014).

Given the citation peak publications characteristic and the importance of recent research on CE and big data/IoT, we did not use citations over time windows analysis in this study.

We also used greenhouse gas emissions and gross domestic product (GDP) data from the World Bank to generate proportional comparisons with publications.

Results were obtained with the use of R statistical software (R Core Team 2016), and R additional packages (Feinerer et al. 2008; Bouchet-Valat 2014; Wickham 2009; Fellows 2014; Hornik et al. 2009; Csardi and Nepusz 2006).

Findings and results

Publication numbers for CE

This section shows the analysis on publications concerning CE. Figure 4 shows the number of publications between 2006 and 2015 grouped by region, all showing a growing interest on the subject. Europe leads with a big gap, while Asia demonstrates more concern for CE than North America in recent years. We observed a 219% growth in publications from 2011/2015 against 2006/2010. In South America the growth was of 329%, followed by the Middle East with 312%, both higher than other regions, which indicates the concern on CE in Asia and North America, for example, is more consistent. Given that circular economy initiatives and researches have become more intense since the creation of the Ellen MacArthur Foundation in 2010, this contributes and justifies the higher publications discrepancy in the past five years. Nevertheless, the 37% publication growth from 2012 to 2015 also confirms there is a growing interest on the subject.

Fig. 4
figure 4

Publication profile on CE by region 2006–2015

Full size image

The bibliometric descriptive results for 80% of all publications on CE from 2006 to 2015, along with a comparison on greenhouse gas emissions, summarized by country, is presented in Table 1. United States is the country with more publications, and also has the most cited paper, although almost half of the country’s papers (46.94%) don’t have a single citation. China follows with a gap of nearly 2000 publications. Switzerland and Sweden are the countries with the highest publications/greenhouse gas emissions ratio (7.5 and 6.9). Although both countries together don’t have a high number of publications (behind Canada), greenhouse gas emissions are low, showing a genuine concern with the subject. At the other end, Brazil and China have the lowest ratio (0.21 and 0.28), indicating more scientific studies need to be conducted and applied in these countries in order to produce more tangible results.

Table 1 Countries representing 80% of all publications on CE compared to greenhouse gas emissions—2006–2015
Full size table

Regarding CPP, considering the growth tendency for publications on CE in recent years (Fig. 4), for this study this index was used as a reference for comparison in country level only. As the citation peak occurs around the 4th publication year, it is expected that for years 2012–2015 the number of citations might increase significantly. In order to provide a method for binding publications to a specific region and avoid overlapping in cases of collaboration among countries from separate regions, for each publication the country/region of the main author’s affiliated institute was considered. This applies to all charts and tables of this research.

We have also developed the social network analysisFootnote 6 for CE publications for the year 2015 in order to understand the current research trends in terms of keywords associations. As we can see in Fig. 5, the terms “Life Cycle Assessment” and “Renewable Energy” are the nodes with higher betweenness centralities, as they form the densest bridges with other nodes. The word China is directly connected to CE and the central node renewable energy, confirming the high country’s concern with the subject, as no other country name is part of the network.

Fig. 5
figure 5

Social network analysis for CE in 2015—Fruchterman-reingold lay-out

Full size image

Publication numbers for CE with big data/IoT

The combination of CE and big data/IoT queries resulted in 70 documentsFootnote 7 that discuss both subjects, as shown on Fig. 6.

Fig. 6
figure 6

CE and big data/IoT docs search summary 2006–2015

Full size image

By filtering the CE results on big data/IoT (Fig. 7), we observed similar trends for Europe, Asia and North America. No publications for Africa, Central America & Caribbean, Middle East, Oceania or South America were found.

Fig. 7
figure 7

Publication profile on CE and big data/IoT by region 2007–2015 (no data for year 2006)

Full size image

Although the use of cutting edge technologies is considered an advantage for CE initiatives, only 0.23% of the published studies considered the use of big data and/or IoT technologies. Nevertheless, a growing interest on the subject has been observed on the last three years.

Table 2 contains the bibliometric descriptive results for all publications on CE along with big data/IoT from 2006 to 2015, along with a comparison on greenhouse gas emissions (GHGE) and gross domestic product (GDP), summarized by country. 23 countries published studies on the subject. China leads in publications, although 74% of their studies have not been cited yet. 52% of all publications are concentrated in China, United States and Germany, and 61% of the countries have only one single publication. We noticed the presence of smaller economies like Romania, Pakistan and Kazakhstan and the absence of big economies, such as Brazil, Japan and Russia. The use of GHGE and GDP data helped confirm the strong relation between GDP and emissions, but not necessarily with publications. For example, Serbia, Finland and Greece appear with better publications/GDP ratios, while France, United States, and Canada show poor results. Also, Pakistan produces more greenhouse gases than Finland, but have fewer publications. India, Canada, United States and China have the lowest publication/emissions ratios, meaning stronger investments in technology for Circular Economy could have been done by those countries.

Table 2 Countries with publications on CE and big data/IoT compared to greenhouse gas emissions—2007–2015
Full size table

We also extracted data at institution level. By analyzing the results in Table 3, we noticed a Finnish university is the most cited institution on the subject, although it produced only two publications. China is the country with most institutions (four), followed by USA (three) and Finland, Italy, UK with two institutions each. The complete institutions list is on Appendix 3.

Table 3 Institutions with cited publications on CE and big data/IoT—2006–2015
Full size table

Content and social network analysis: CE with big data/IoT

The corresponding articlesFootnote 8 were submitted to content analysis throughout the entire articles contents (references excluded). After documents cleansing (prepositions, numbers, punctuation, common words), we came up with a list of the most frequent words (word-stem), detailed on Fig. 8, which presents terms that lead to researches more related to the public sector/sponsored by government agencies. The additional omitted common words list is presented on Appendix 4. Main research words were also omitted: data, information, system, big data, sustainability, sustain, circular economy. Words such as environment, service, power, efficiency shown in high positions suggest a higher research focus on these areas.

Fig. 8
figure 8

Word cloud with top 150 terms (word-stem) found on available articles

Full size image

Besides the high incidence of common terms, the identification of some other relevant keywords indicate current researches occurring in areas such as carbon emissions reduction, smart cities, water economy, environment, waste management and renewable sources as well. Cloud computing has been an additional enabler.

The social network analysis (Fig. 9) based on all articles’ keyword fields shows some new terms highly central and connected nodes with betweenness centralities, such as “Open Data”, “Energy Efficience”, “Environment” in one network, and a balance among terms on the other network. The high frequency and connections of the term “Sustainable HCI” (human–computer interaction) deserves to be pointed out as an interesting finding, along with “ICT Research Strategy”, showing a real interest for information and technologies research on the area, not necessarily bound to big data/IoT.

Fig. 9
figure 9

Social network analysis for CE with big data/IoT—Kamada—Kawai layout

Full size image

We depicted the cited papers on CE with the use of big data/internet of things on Table 4, along with their SNA-related keywords. Some of the researches focus on initiatives possibly related in a way with government agencies and the public sector in general, given the regular use of terms such as smart grid, sustainable cities, utilities, and environmental politics, also present on the word cloud analysis. Other interesting finding is the variety of applications encountered: closed loop product lifecycle management, establishment of sustainable cities, energy grid optimization, ecologically designed products and buildings, sustainable agriculture, for example, meaning the use of the concepts applies to almost every action we take in our personal or business lives.

Table 4 Cited papers on circular economy with the use of big data/internet of things
Full size table

We also selected the articles with more than ten citations from the 70 papers found to explore in more detail the researches being undertaken in the area of CE with the support of big data/IoT:

The use of such technologies on the development/leverage of CE initiatives can be found in a variety of environments and has several different applications. One is called sustainable PLM (product lifecycle management), which focuses on the use of IoT to make devices (called intelligent products) to interact among themselves in order to promote environment benefits, such as energy usage optimization, CO2 emissions/environment impacts reduction (Främling et al. 2013). One real example comes from the automotive industry: sensors used on vehicles engines communicate to the vendors via Bluetooth to mobile phones, which send the data to car service provider companies/manufacturers. These provide online and continuous diagnostics, identify engine error conditions that lead to fuel overconsumption and communicate it to the owner, allowing predictive maintenance, spare parts/mechanics availability and, mostly important, promote fuel optimization and vehicle life cycle extension. Smart houses and appliances are also an example of the same application (Främling et al. 2013). Extended applications of similar technologies can also be used by industries, enhancing its benefits for the environment. Some electric power generator companies, for example, use big data technologies to support identifying the location of their wind turbines in order to improve the efficiency of electric power generation, or the big data technology to better balance and integrate the power generation from new or renewable resources. Real-time monitoring, enabled by big data technologies, also improve the automation level of the power systems where it has already been implemented (Peng et al. 2015).

IoT and big data have also been playing a fundamental role on the development of smart cities. The concept of environmental internet of things (EIoT), for example is being studied in China, due to the country’s population and pollution growth. By implementing and integrating sensors for mobile (e.g. aircrafts, boats, cars, cell phones) and stationary assets, the technology can monitor soil, water, wind etc. providing on-line and real-time environment information, and supports decision-making processes for the management of future sustainable cities (Zhao et al. 2013). It is also recognized that smart city concepts are still immature and will require a lot of planning and standardization in order to allow different infrastructures, subsets and devices to properly communicate and interact in an acceptable manner. Some researchers are already trying to come up with a smart city model to be used as a repeatable and exportable standard, such as the Intelligent Distributed Autonomous Smart City—IDASC, which is a conceptual model that embraces diversity of technology, operators, and connection (Roscia et al. 2013).

The consumer electronics industry is also taking advantage from the benefits CE leveraged by technology. The recent advances in electronic technology has shortened the life cycle of components, leading to the growth of electronic and electric waste (e-waste), which has been raising concerns about the environment. Ecodesign models are now being applied to such components, resulting in the development of more environmentally friendly materials. And IoT has a crucial role on the success of e-waste reduction, as consumer electronics logistics in all supply chain network (including material selection, product distribution and e-waste collection) tends to be improved through equipping components with RFID tags and other types of sensors already available at low cost. The use of supporting databases, geolocation/decision support/integrated environmental information systems will provide a huge contribution to e-waste management, saving resources from materials use (Li et al. 2015).

We should also mention some key CE terms not shown on our big-data/IoT research. Those indicate studies opportunities are even broader for researchers: bioeconomy, carbon biofixation, carbon capture, carbon recovery, climate bond, circular ecosystem, eco industry, end of waste, sharing cities, waste recovery, waste to value.

Policy implications

There are some ways in which our findings would support a move towards more motivation in researches on CE with big data/IoT, and, as a consequence, more actions on moving from the linear to the circular model, leveraged by technology. First, the gap found between the types of current scientific researches on the field and corporate initiatives can be reduced by the time they work closer, in a complimentary way, as both sides will benefit from such initiatives. This may then act as an incentive to make them work more closely to each other (by making action researches, for example). Second, as the keywords search queries presented in the methodology section were obtained empirically due to the lack of predefined taxonomies for CE, they can in future researches be organized and transformed in structured taxonomies, turning the scientific exploration of CE less complicated. Also and consequently, as the subject of this research is relatively new, some papers might not have been part of the query results, as the respective authors might not have used the selected keywords.

Moreover, the findings can also serve as an alert to those countries identified as with a high rate greenhouse gas emissions compared to scientific studies, so they can be motivated to start moving towards a circular approach.

In addition, by understanding the studies being undertaken in the field, we can suggest a key orientation to organizations, which is to start thinking about CE from the design phase of their products, as it is a concept present through all chains on the production system. By designing products with modular components, for an extended life-cycle, reusable, trackable, easily disassembled and made from renewable sources rather than only being concerned about the product’s visual aspect, this will be far more acceptable by consumers (which are becoming more and with more environmental awareness) than using only a visual appealing product that harms the environment.

Conclusions

Undoubtedly, the attractiveness of CE is growing globally and changes in consumer behavior are playing an active role in this transformation, fostering researches all over the world. The application of technologies such as big data and IoT, being also enablers, has been showing a consistent growth interest on recent years but still requires more studies in the field. Some large economies in terms of geographic area, economy and greenhouse gas emissions (therefore with potential interest on technology applied to CE) such as Brazil (along with the rest of South America) and Russia are still not generating any publications in the field. Specifically in Brazil, we noticed a growing interest on the subject leveraged by independent entrepreneurs, which are collaborating with successful countries and fostering local initiatives, such as the Exchange 4 ChangeFootnote 9 organization.

Some fundamental principles belonging to the CE framework require attention, as they are still not being scientifically studied globally with the support of technology, such as bioeconomy and waste recovery.

Moreover, as most scientific studies found on this research consist on conceptual models, ongoing projects, prospects, tendencies, pilots, they are more now like “imagining the possibilities” than actually developing cases studies based on already established programs for benefits measuring. On the other hand, there are already interesting industry/private sector cases waiting to be scientifically explored by researchers, such as the examples illustrated in this document.

Additionally, the study can help researchers develop more detailed and specific CE bibliometric analysis to areas other than big data/IoT with the use of the preliminary keyword taxonomy mapped in this research.