3.1 Systematic literature review
According to Sampaio and Mancini (2007), a systematic review requires a clear research question, the definition of search strings, the establishment of inclusion and exclusion criteria for articles, and an analysis of the selected literature. The databases for consultation should be defined, as well as the criteria for conducting searches and selecting relevant materials, ending with data extraction and a critical analysis of these.
Based on the understanding of systematic literature reviews and the objective of this research, search strings were developed and tested using boolean operators AND, OR, and AND NOT to construct advanced search strategies, where AND corresponds to conjunction, OR corresponds to the union, and AND NOT corresponds to exclusion (Galvão & Ricarte, 2019), for application in the selected databases, namely Emerald, Taylor and Francis, Springer, Wiley, Scopus, Web of Science, and ScienceDirect.
After the initial tests, the string to be used was defined: ("Design for Reverse Logistics" OR ("Design for X" AND "Reverse Logistics")) AND NOT "Network Design" AND NOT Network. This was applied to the considered databases. In the next step, the results were analysed according to previously defined inclusion and exclusion criteria (Table 1).
Inclusion criteria
|
- Contain related keywords in the papers' title, abstract or keywords;
- Papers published within the last 20 years;
- Papers written in Portuguese or English.
|
Exclusion Criteria
|
- Trabalhos que não contenham nenhuma das palavras-chave;
- Trabalhos hospedados em plataformas pagas, não custeadas pela UEPA;
- Capítulos de livro;
- Trabalhos que não apresentem pelo menos 2duas palavras chaves aplicadas na string relacionadas.
- Papers that do not contain any of the keywords;
- Papers hosted on paid platforms not sponsored by the university (i.e. UEPA);
- Book chapters;
- Papers that do not present at least 2 applied keywords in the related string.
|
Table 1: Inclusion and Exclusion Criteria
After analysis based on the criteria presented, articles were selected for a more thorough analysis, already generating some results for the present research. In the Scopus database, a significant presence of works on RL was noticed, but not focused on the main theme of this research. Even with the application of the "AND NOT" operator to exclude works on "Networking design," many articles with this theme still appeared. Other works presented "reverse logistics" in the abstract or keywords only brought this term as a supporting concept or only superficially mentioned it.
During the research conducted in the Scopus and ScienceDirect databases, many papers were found that solely focused on supply chain and waste reverse channels with the potential for energy generation. However, some interesting works were identified that addressed the integration of CE into Product Design. These studies also considered Life Cycle Assessment along with CE, highlighting that studying end-of-life products during the design stage can lead to the creation of viable products for the CE reality (Mestre and Cooper 2017; Fofou et al. 2021; Hapuwatte and Jawahir 2021; Wang et al. 2022) and for Reverse Logistics design (Chouinard et al. 2009; Daaboul et al. 2016).
After the analysis of the found works through inclusion and exclusion criteria, 26 articles were selected for data extraction. It was noticed that most of the research is based on the need to rethink human consumption, reducing the extraction of raw materials and focusing on options for recovering waste that is currently only being discarded, harming the environment. This reality causes several problems, such as the increasing improper disposal that accompanies population growth (Marrucci et al. 2019). Research on RL already exposes the demand for more studies that enable its establishment (Sharma and Singh 2013; Huscroft et al. 2013; Wilson and Goffnett 2022).
In their research aimed at exploring and understanding how Design can contribute to a transition to CE through the adoption of DfX approaches, Sassanelli et al. (2020) add the concept of servitization to the DfX and CE concepts, indicating the trend and need for reinvention of more sustainable processes, products, and services. The authors propose to analyse possible approaches between DfX and CE. However, they also assert that even though various DfX approaches related to end-of-life stages have already been proposed (e.g. Design for Recycling, Design for Remanufacturing, and Design for End-of-life), they still need to be better aligned with the perspective of Circular Design.
Bringing terms such as Circularity and proposing life cycle analysis to identify interventions, or proposing methodology for the design of circular products, Wilson and Goffnett (2022) argue that a major obstacle encountered in these research is the issue of RL. It was also noted that some researchers have recently been trying to contribute to the feasibility of RL, as in Wilson, Paschen, and Pitt (2022), where an integration between CE and Artificial Intelligence was proposed to enable RL processes. Another example is the research by Julianelli et al. (2020), who criticize the production and consumption practices that follow the "take-make-dispose" flow and their negative impacts, leading society to seek sustainable development options, highlighting CE as a relevant concept that promotes the RL for the end-of-life products.
It was also found that most of the works that integrate circularity, green chains, CE, or RL concepts involve Design being associated with these concepts (den Hollander et al. 2017; Molla et al. 2022; Harun et al. 2022; Relich et al. 2022; Mallick et al. 2023; Piao et al. 2023; Burke et al. 2023). This makes it clear that, given the problems that humanity is currently facing, designers must be familiar with every phase of a product's life cycle in order to rethink their impacts (Papanek 1984).
According to the information found, it is possible to see that an integration between the DfX approaches and RL - which would be DfRL - has great potential to enable RL from product conception, and consequently enable circular concepts found, such as CE and Circular Design.
3.2 Identification and validation of design needs
The results of the literature review served as a basis for initiating discussions within the Focus Group dynamics. The RL activities were presented, as well as their concepts and other information contained in Melo et al. (2022) research. These specialists were asked to identify the DN in each of these activities. The result of the dynamics is presented in Table 2.
RL Activities
|
Characterization of RL Activities
|
DN identified in the Focus Group dynamics
|
Integration
|
This is the first and most important link in the RL process. It is the planning stage that occurs throughout the entire process. It also seeks to integrate the entire RL process through information sharing among all members of the reverse channel.
|
Have access to information such as manufacturers, resellers, and potential customers (types), indication of recovery points (recoverable waste), and proper disposal (non-recoverable waste);
|
Waste Acquisition
|
In this activity, based on some information collected in the Integration stage, the generation sources are identified, the types of waste and the quantities to be collected are defined.
|
*** No DN were identifyed;
|
Gatekeeping
|
This refers to the refined selection of waste. It begins when the materials to be collected must be sorted to enter the return flow according to their functionality and to determine the flow to be followed in the RL process.
|
Generate clear/universal information about recommended recovery materials and methods;
|
Collection
|
This activity starts with the transfer of materials to be returned from the waste generators to processing centers. This transfer can occur passively, when collection agents go to the generation point, or actively when those responsible for generation make some type of movement so that the transfer occurs, or it can be combined.
|
Sensitize the user that the waste can be collected or delivered for recovery after its use.;
|
Transport
|
This is an activity exclusively related to the movement of material (transportation, loading/unloading, handling) between facilities or activities in reverse channels.
|
Promote stability, reduction of clearances; Optimization of cargo space in vehicles; Easy handling; Easy disassembly; Easy stacking; Easy compaction;
|
Inspection/
Testing
|
This is a more detailed investigation of waste, considering technology and materials, general condition and functioning, aiming to evaluate the most sustainable recovery alternative. Depending on the waste characteristics, it should range from observing the general condition and a simple on/off test to more comprehensive testing.
|
Facilitate the diagnosis of the general condition of the waste; Facilitate the evaluation of possibilities for waste recovery; Provide an evaluation protocol of the conditions and operations of the waste, developed by the manufacturer;
|
Disassembly
|
This is characterized by the organized dismantling/disassembly of partial or total waste into portions, pieces, subparts, or basic components.
|
Promote more fits than screws; Provide techniques designed not to damage the product and its components during disassembly;
|
Sorting
|
In this activity, the most appropriate recovery and disposal alternatives are defined, considering certain associated criteria. Sufficient information should be generated to support decisions regarding the fate of waste (parts, subparts, basic components, or materials) for the most appropriate recovery options or proper ways of disposal.
|
*** No DN were identifyed;
|
Redistribution
|
Consists of two stages. The first encompasses all activities related to the effective sending of processed materials from processing points to appropriate recovery units or disposal facilities. The second stage of this activity aims to reintegrate directly reusable products and recovered materials or products into the most suitable production cycles (industry, commerce, and services) along forward supply chains.
|
*** No DN were identifyed;
|
Warehousing
|
This activity refers to all actions aimed at maintaining the current state of conservation of the waste to be recovered or the new state of consumption of the already recovered waste. It includes all material handling and internal movement activities in the facilities that make up the reverse channels.
|
Be stackable
|
Table 2: Characterization of the activities of RL and DN pointed out during the Focus Group dynamics. Developed by the authors based on Melo et al. (2022).
The suggested DN by the Focus Group were analysed in light of the materials used as a basis, and based on this analysis, some of them were relocated, and a final list of DN was elaborated. Some demands from the RL activities resulted in more than one DN in the final list, totalling 21. Table 3 relates these DN to the associated RL activities.
Code
|
Design Needs (DN)
|
RL Activities
|
DN01
|
Make available in the product information such as: manufacturer; retailer(s) and possible customers (types); indication of recovery points (recoverable waste); and proper disposal (irrecoverable waste).
|
Integration
|
DN02
|
Provide information about the volume and type of waste generated (or present) at each point of generation, for potential users of these wastes.
|
DN03
|
Provide information about the location of waste generation points.
|
Waste Acquisition
|
DN04
|
Provide information about the identification of the product/component/waste, to assist in the decision-making process regarding its recovery or disposal.
|
Gatekeeping
|
DN05
|
Provide information on proper procedures for preparing the product/component/waste for its recovery (e.g. cleaning, densification, disassembly, etc.).
|
DN06
|
Have artifacts available to facilitate the collection of discarded products/components/waste (e.g. waste collection containers).
|
Collection
|
DN07
|
Spread information to society about voluntary delivery points and waste disposal points.
|
DN08
|
Raise awareness among users that products/components/waste can be collected (or delivered) for recovery and when this collection takes place.
|
DN09
|
Have a heterogeneous and flexible fleet of vehicles available for the transportation of collected products/components/waste.
|
Transport
|
DN10
|
Optimize the space occupied by the product/component/waste in the transport vehicle (e.g. disassembly, stacking, and compression).
|
DN11
|
Preserve the integrity of the product/component/waste during transportation (e.g. stability, reduction of clearances, etc.).
|
DN12
|
Provide a protocol for evaluation, preferably developed by the manufacturer, regarding the state of integrity (conservation and operation) of the product/component/waste.
|
Inspection/
Testing
|
DN13
|
Use fittings as elements for joining parts of a product/component/waste, instead of screws and welds.
|
Disassembly
|
DN14
|
Provide information on the disassembly process of the product/component/waste (e.g. website or video with step-by-step description).
|
DN15
|
Provide a protocol for classifying the level of integrity (conservation and operation) of the product/components, preferably developed by the manufacturer or qualified stakeholder, to perform the recovery.
|
Sorting
|
DN16
|
Preserve the integrity of the product/component/waste (e.g. stability, reduction of clearances, packaging, etc.) during its shipment both for recovery (materials to be recovered) and for its reintegration into the supply chain (materials already recovered).
|
Redistribution
|
DN17
|
Facilitate handling (packaging, palletizing, etc.) of the product/component/waste throughout the entire process.
|
Warehousing
|
DN18
|
Enable compression of the product/component/waste (e.g. bundling, stacking, etc.).
|
DN19
|
Have a diversified and flexible storage system (structures and environmental characteristics) available for the product/component/waste throughout the entire process.
|
DN20
|
Keep diversified and flexible handling equipment (forklifts, pallet jacks, carts, etc.) available for the product/component/waste handling throughout the entire process.
|
DN21
|
Maintain inventory control of products/components/waste (quantities of inputs and outputs, varieties and, if applicable, storage deadlines) throughout the entire process.
|
Table 3: Design needs based on reverse logistics activities according to this research.
From the final list of DN associated with each activity of the RL, resulting from the literature review and application of the Focus Group method, a questionnaire was developed for the application and the collection of opinions (survey) from specialists on how each proposed DN is decisive for the RL activities.
To create a comprehensive survey for this research, we engaged with three groups of specialists who were carefully selected based on their expertise and experience in relevant fields. The groups included professionals specializing in reverse logistics, sustainable design, and circular economy. These specialists were chosen for their specific knowledge and insights related to the topic of our research. We then distributed the questionnaire to these groups to gather a thorough and insightful understanding of the subject matter. Circular economy specialists were included because they are multidisciplinary professionals and have a cyclic perspective at their core (Ellen MacArthur Foundation 2015). The questionnaire was sent via email, and the specialists' email addresses were obtained through the Lattes Platform or LinkedIn.
The questionnaire was sent to a total of 496 specialists. The current total number of respondents was 81, which represents 16.33% of the total sample. The questionnaire includes a space for contributions from respondents, where, if they wish, they can provide some reflection or point out a DN for RL that has not been considered until then. As a result, there have already been significant contributions, in addition to qualitative responses, for the stage of analysing results, insights, and gaps for future research on the topic.
Upon obtaining the survey database, the treatment initially occurred through the calculation of Cronbach's Alpha (𝛼), which is a statistical tool used to analyse the reliability of the questionnaire so that it is appropriately designed and reliably reproduces the reality of the study (Ribeiro et al. 2021). It should be noted that for this stage, the guidelines of Christmann and Van Aelst (2006) were followed, which present 0.70 as the minimum acceptable value, validating and ensuring the reliability of the questionnaire used. For Freitas and Rodrigues (2005), this value would be at least 0.60, and below this value, the internal consistency of the scale used is considered low. In this research, the obtained 𝛼 was 0.9113, which ensures the reliability of the questionnaire used.
3.3 Hierarchization of design needs
The multi-criteria analysis method defined for application in this research was TOPSIS, primarily because it provides a simpler data collection process and requires fewer judgments. Additionally, it is considered suitable for modelling values of known quantitative criteria accurately (Lima Junior and Carpinetti 2015).
The TOPSIS algorithm aims to evaluate a decision matrix containing all viable alternatives associated with attributes and weights, in order to rank the alternatives from the most similar to the ideal solution to the least similar (Hwang and Yoon 1981). Thus, the method generates a descending order, where the alternative that presents the highest relative proximity will be ranked first. After the steps of matrix construction and calculations that are part of the method, using alternatives and weights, the alternatives are sorted, starting from the similarity values calculated for each alternative, they must be ordered in a decreasing way, where results close to 1 are preferable to results close to 0 (Freitas and Costa 2018). Therefore, it is possible to affirm that the application of the method follows the same logical structure of a decision-making process carried out by a decision maker to order alternatives.
According to Singh et al. (2016), TOPSIS allows for the classification of alternatives considering different criteria of analysis (which can be weighted differently). This is the case in the present research, where it is understood that the specialists' experience time in their respective expertise is a factor that can interfere in their experiences to evaluate the considered DN. The methodological procedure described by Singh et al. (2016) exposes that initially a Matrix D (Equation 1), with elements xij, where i represents an alternative and j a criterion of analysis, must be structured.
![](https://myfiles.space/user_files/122228_c8a1650c59388082/122228_custom_files/img1697007192.png)
Next, the matrix D must be normalized and the coefficients rij must be calculated. Based on these, they should be weighted with the weights established by the researchers and then calculate the values vij (Equation 2), which will make up the Matrix V (Equation 3).
![](https://myfiles.space/user_files/122228_c8a1650c59388082/122228_custom_files/img1697007268.png)
![](https://myfiles.space/user_files/122228_c8a1650c59388082/122228_custom_files/img1697007284.png)
From the matrix V, it is necessary to calculate the positive ideal solution (vj+) and the negative ideal solution (vj-) using the equations presented in Equation 4. With this data, it is possible to calculate the positive and negative Euclidean distances for each alternative.
After obtaining the positive and negative Euclidean distances for each alternative, the Ci* indicator is calculated. The higher the values of Ci*, the better the results. These values range from 0 to 1. Equation 5 shows how this indicator is calculated. Finally, the alternatives should be ordered according to the Ci* values that were calculated.
![](https://myfiles.space/user_files/122228_c8a1650c59388082/122228_custom_files/img1697007329.png)
The development of the TOPSIS method was carried out as presented. All data collected through the questionnaire were applied to a structured spreadsheet, according to the steps described by Singh et al. (2016). In this research, in order to hierarchize the DN for RL, weights were first assigned to the specialists' responses based on their years of experience. These values are presented in Table 4. It is noteworthy that Group 1 corresponds to specialists with more than 20 years of experience, Group 2 to specialists with between 10 and 20 years of experience, and Group 3 to specialists with up to 10 years of experience.
Weight
|
Group 1
(more than 20 years)
|
Group 2
(between 10 and 20 years)
|
Group 3
(up to 10 years)
|
0,50
|
0,30
|
0,20
|
Table 4: Weights assigned to the specialists.
Table 5 presents the average ratings assigned by the professional specialists to each of the DN presented in Table 3.
Code
|
Group 1
|
Group 2
|
Group 3
|
DN01
|
4,66
|
4,50
|
4,51
|
DN02
|
4,41
|
3,86
|
4,34
|
DN03
|
4,25
|
3,90
|
4,05
|
DN04
|
4,58
|
4,13
|
4,48
|
DN05
|
4,66
|
4,00
|
4,57
|
DN06
|
4,91
|
4,45
|
4,80
|
DN07
|
4,91
|
4,77
|
4,68
|
DN08
|
4,91
|
4,45
|
4,85
|
DN09
|
4,25
|
3,63
|
3,82
|
DN10
|
4,25
|
3,63
|
4,11
|
DN11
|
4,50
|
3,45
|
4,25
|
DN12
|
3,91
|
3,63
|
3,97
|
DN13
|
4,16
|
3,22
|
4,08
|
DN14
|
3,66
|
3,40
|
4,34
|
DN15
|
3,58
|
3,50
|
4,25
|
DN16
|
4,16
|
3,72
|
4,31
|
DN17
|
4,16
|
4,27
|
4,37
|
DN18
|
4,08
|
4,09
|
4,37
|
DN19
|
3,83
|
3,72
|
4,05
|
DN20
|
4,08
|
3,50
|
3,80
|
DN21
|
4,33
|
3,77
|
4,22
|
Table 5: Means attributed by each group of specialist respondents to each of the DN for RL.
With the means attributed by each group of specialist respondents to each of the DN for RL, we proceeded to their ranking. Using the means presented in Table 5, we performed normalization of the values, resulting in the Matrix R, presented in Table 6.
Code
|
rij (more than 20 years)
|
rij (between 10 and 20 years)
|
rij (up to 10 years)
|
DN01
|
0,24
|
0,25
|
0,23
|
DN02
|
0,22
|
0,22
|
0,22
|
DN03
|
0,21
|
0,22
|
0,21
|
DN04
|
0,23
|
0,23
|
0,23
|
DN05
|
0,24
|
0,22
|
0,23
|
DN06
|
0,25
|
0,25
|
0,24
|
DN07
|
0,25
|
0,27
|
0,24
|
DN08
|
0,25
|
0,25
|
0,25
|
DN09
|
0,21
|
0,20
|
0,19
|
DN10
|
0,21
|
0,20
|
0,21
|
DN11
|
0,23
|
0,19
|
0,22
|
DN12
|
0,20
|
0,20
|
0,20
|
DN13
|
0,21
|
0,18
|
0,21
|
DN14
|
0,19
|
0,19
|
0,22
|
DN15
|
0,18
|
0,20
|
0,22
|
DN16
|
0,21
|
0,21
|
0,22
|
DN17
|
0,21
|
0,24
|
0,22
|
DN18
|
0,21
|
0,23
|
0,22
|
DN19
|
0,19
|
0,21
|
0,21
|
DN20
|
0,21
|
0,20
|
0,19
|
DN21
|
0,22
|
0,21
|
0,21
|
Table 6: R Matrix with normalized values of DN for RL
Later, weights were assigned to each group of specialist respondents according to the information presented in Table 4. Using this procedure, it was possible to obtain the V Matrix, as shown in the values presented in Table 7.
Code
|
rij (more than 20 years)*0,50
|
rij (between 10 and 20 years)*0,30
|
rij (up to 10 years)*0,20
|
DN01
|
0,12
|
0,08
|
0,05
|
DN02
|
0,11
|
0,06
|
0,04
|
DN03
|
0,11
|
0,07
|
0,04
|
DN04
|
0,12
|
0,07
|
0,05
|
DN05
|
0,12
|
0,07
|
0,05
|
DN06
|
0,12
|
0,07
|
0,05
|
DN07
|
0,12
|
0,08
|
0,05
|
DN08
|
0,12
|
0,07
|
0,05
|
DN09
|
0,11
|
0,06
|
0,04
|
DN10
|
0,11
|
0,06
|
0,04
|
DN11
|
0,11
|
0,06
|
0,04
|
DN12
|
0,10
|
0,06
|
0,04
|
DN13
|
0,11
|
0,05
|
0,04
|
DN14
|
0,09
|
0,06
|
0,04
|
DN15
|
0,09
|
0,06
|
0,04
|
DN16
|
0,11
|
0,06
|
0,04
|
DN17
|
0,11
|
0,07
|
0,04
|
DN18
|
0,10
|
0,07
|
0,04
|
DN19
|
0,10
|
0,06
|
0,04
|
DN20
|
0,10
|
0,06
|
0,04
|
DN21
|
0,11
|
0,06
|
0,04
|
Table 7: V Matrix with weighted values of DN for RL
Next, Table 8 shows the positive and negative ideal solutions. Such information is necessary to calculate the values that correspond to the Euclidean distances from the positive and negative ideal solutions. Then, it was possible to calculate the Ci* coefficient, which was used for ranking. These results are presented in Table 9.
Criterion of solution
|
Specialists time of experience
|
More than 20 years
|
Between 10 and 20 years
|
Up to 10 years
|
Positive ideal solution (vj+)
|
0,12
|
0,08
|
0,05
|
Negative ideal solution (vj-)
|
0,11
|
0,06
|
0,04
|
Table 8: Positive ideal solution and negative ideal solution for accessing the criteria.
Code
|
Euclidean distances of the positive ideal solution (Si+)
|
Euclidean distances of the negative ideal solution (Si-)
|
Coefficient (Ci*)
|
DN01
|
0,01
|
0,04
|
0,81
|
DN02
|
0,02
|
0,02
|
0,54
|
DN03
|
0,02
|
0,02
|
0,46
|
DN04
|
0,01
|
0,03
|
0,68
|
DN05
|
0,01
|
0,03
|
0,68
|
DN06
|
0,01
|
0,04
|
0,88
|
DN07
|
0,00
|
0,04
|
0,96
|
DN08
|
0,01
|
0,04
|
0,88
|
DN09
|
0,03
|
0,02
|
0,40
|
DN10
|
0,03
|
0,02
|
0,41
|
DN11
|
0,03
|
0,02
|
0,49
|
DN12
|
0,03
|
0,01
|
0,25
|
DN13
|
0,03
|
0,02
|
0,31
|
DN14
|
0,04
|
0,01
|
0,14
|
DN15
|
0,04
|
0,01
|
0,14
|
DN16
|
0,03
|
0,02
|
0,40
|
DN17
|
0,02
|
0,02
|
0,53
|
DN18
|
0,02
|
0,02
|
0,45
|
DN19
|
0,03
|
0,01
|
0,24
|
DN20
|
0,03
|
0,01
|
0,30
|
DN21
|
0,02
|
0,02
|
0,48
|
Table 9: Euclidean distances of the positive ideal solution, Euclidean distances of the negative ideal solution, and Ci* Coefficient
Finally, by organizing the obtained values of Ci* coefficient, a comparative ranking of the DN for RL is obtained considering which ones would be more or less decisive to enable RL, according to the point of view of the consulted groups of specialists. Table 10 presents the results of the aforementioned ranking considering its association with RL activities.
Ranking
|
DN Code
|
Ci*
|
RL Activity
|
1º
|
DN07
|
0,962
|
Collection
|
2º
|
DN08
|
0,885
|
3º
|
DN06
|
0,884
|
4º
|
DN01
|
0,806
|
Integration
|
5º
|
DN04
|
0,683
|
Gatekeeping
|
6º
|
DN05
|
0,680
|
7º
|
DN02
|
0,542
|
Integration
|
8º
|
DN17
|
0,526
|
Warehousing
|
9º
|
DN11
|
0,487
|
Transport
|
10º
|
DN21
|
0,481
|
Warehousing
|
11º
|
DN03
|
0,465
|
Waste Acquisition
|
12º
|
DN18
|
0,451
|
Warehousing
|
13º
|
DN10
|
0,411
|
Transport
|
14º
|
DN16
|
0,402
|
Redistribution
|
15º
|
DN09
|
0,398
|
Transport
|
16º
|
DN13
|
0,313
|
Disassembly
|
17º
|
DN20
|
0,297
|
Warehousing
|
18º
|
DN12
|
0,251
|
Inspection e testing
|
19º
|
DN19
|
0,244
|
Warehousing
|
20º
|
DN14
|
0,144
|
Disassembly
|
21º
|
DN15
|
0,139
|
Sorting
|
Table 10: Design needs (DN) ranking and reverse logistics (RL) activities
3.4 Analysis
From the hierarchy of DN for RL generated by the support of TOPSIS, it was possible to see:
- The existence of 4 hierarchical groups that present relatively close Ci* values that could be grouped;
- It can be observed that the criteria related to the Collection activity, ranked at the top of the list (DN07, DN08, and DN06, respectively), are related to the end consumer, highlighting their participation as a basic element for the entire RL process to occur. This finding confirms what is already stated in the literature (Chaves, Giuriatto, and Ferreira 2020; Melo et al. 2022);
- DN01, which ranks fourth and is part of the first group with higher Ci* coefficients, demonstrates that sharing information about the product with various stakeholders is considered very important, probably because it is a DN that is linked to a fundamental step for the other activities of the RL process to be viable (Chaves et al. 2020);
- The DN06, ranked 3rd, highlights the importance of considering and prioritizing the availability of artifacts that facilitate the collection of post-consumer products, components, and waste, such as collection containers;
- DN04 and DN05 are ranked closely in 5th and 6th place, respectively, and both are related to the Gatekeeping activity, further highlighting the importance of providing product information for the entire process of RL to be viable (Pyke and Tang 2010);
- DN01, DN04, DN05, and DN02 are ranked sequentially and are related to informational activities of RL (Integration and Gatekeeping), such as instructions for correct handling of materials or proper disposal. This grouping may indicate a correlation between these DN;
- The DN related to the warehousing activity are scattered throughout the ranking, present in the second, third, and fourth ranking groups. It is worth noting that most DN for RL were proposed for this activity;
- The DN related to the Transport activity are also scattered throughout the ranking, but within the same group. Similarly, the DN from the Disassembly activity are in the same configuration;
- It is possible to visualize (Table 11) a new grouping dimension that relates the DN to the design goals associated with the RL activities. In this dimension, it is possible to identify the possibilities of interference of DfRL in the product design, support artifacts for RL, and RL support processes (services, systems, procedures, etc.). In other DN, it was observed that this interference can occur in a combined way, with two or more design goals;
Ranking
|
DN Code
|
Design Goal
|
RL Activity
|
1º
|
DN07
|
Processes
|
Collection
|
2º
|
DN08
|
Processes
|
3º
|
DN06
|
Artifacts
|
4º
|
DN01
|
Products
|
Integration
|
5º
|
DN04
|
Products and Processes
|
Gatekeeping
|
6º
|
DN05
|
Processes
|
7º
|
DN02
|
Processes
|
Integration
|
8º
|
DN17
|
Products
|
Warehousing
|
9º
|
DN11
|
Products and Artifacts
|
Transport
|
10º
|
DN21
|
Processes
|
Warehousing
|
11º
|
DN03
|
Processes
|
Acquisition
|
12º
|
DN18
|
Products and Artifacts
|
Warehousing
|
13º
|
DN10
|
Products and Artifacts
|
Transport
|
14º
|
DN16
|
Artifacts
|
Redistribution
|
15º
|
DN09
|
Artifacts
|
Transport
|
16º
|
DN13
|
Products
|
Disassembly
|
17º
|
DN20
|
Artifacts
|
Warehousing
|
18º
|
DN12
|
Processes
|
Inspection e testing
|
19º
|
DN19
|
Artifacts and Processes
|
Warehousing
|
20º
|
DN14
|
Artifacts
|
Disassembly
|
21º
|
DN15
|
Processes
|
Sorting
|
Table 11: Design Needs (DN) and Design Goals for Reverse Logistics (RL)
- According to Table 11, it is possible to observe that the DN oriented towards processes, either in isolation or in combination, are more frequent (10 occurrences) and more present in the top 10 positions of the table (6 occurrences, 5 of them in isolation). It is believed that this is due to the fact that RL is still a requirement that is not well-considered in current product design and, to make it viable at the present time, it is more effective to intervene in processes due to their intangible nature, which is less expensive to create and modify;
- According to Table 11, it is possible to observe that product-oriented DNs for RL occurred in isolation (3) and in combination (4). Even with a lower overall occurrence frequency (7) compared to process-oriented DNs, it was noted that this design goal was also quite frequent (4 occurrences) among the top 10 positions in Table 11. Such a result may reflect the influence of RL on the physical integrity of products and waste that flow along reverse channels, as poorly executed RL may alter the current state of conservation of the product or waste, leading it to become a non-recoverable waste;
- It is also noted that DN oriented towards RL artifacts were also quite frequent (9 occurrences) throughout Table 11, but their frequency in the top 10 positions was lower (2) compared to other design goal considered. This may have occurred because artifacts developed for forward logistics activities are also used in RL activities. This may explain the lower frequency of DN associated with this design goal in the observed results.
3.5. Contributions to theory and practice
The results of this study contribute to the discussions in the areas of reverse logistics and sustainable design. Specifically, it presents the DN related to the activities that make up the reverse logistics process. A better understanding of these DN allows for definitions to be discussed and solutions to be sought still in the product design stage capable of meeting them. This could make it feasible to execute some reverse logistics activities or increase the efficiency of others that are already being executed. For design, it contributes to starting the debates in defining and evolving the DfRL approach, allowing new discussions on requirements and recommendations for the design of products, processes, and artifacts for reverse logistics to be initiated.
In addition, by seeking to understand the existing DN within the processes that involve and enable the full execution of RL, grounding future measures regarding both academia and industry, this research contributes to achieving SDG 9, specifically target 9.4, which establishes the commitment to "upgrade infrastructure and retrofit industries to make them sustainable, with increased resource-use efficiency and greater adoption of clean and environmentally sound technologies and industrial processes." Similarly, discussions about actions and decisions related to product, artifact, and process design for RL also contribute to achieving target 12.5 (SDG 12), which proposes to "substantially reduce waste generation through prevention, reduction, recycling, and reuse.".