Comparing Metal Additive Manufacturing with Conventional Manufacturing Technologies: Is Metal Additive Manufacturing More Sustainable?

Abstract

CO₂ emissions continue to rise, along with the associated environmental risks. In response, the United Nations has been promoting the adoption of sustainable practices among businesses worldwide. In parallel, an innovative technology known as additive manufacturing (AM) has emerged over the past four decades. This technology has the potential to be more sustainable than conventional manufacturing (CM) technologies. When metals are used as the material, the process is referred to as metal additive manufacturing (mAM). AM technologies have seven process categories, which include metal mAM processes, most notably powder bed fusion (PBF), directed energy deposition (DED), binder jetting (BJT), material extrusion of metal-filled feedstock, and sheet lamination. Among these, PBF and DED are by far the most widely applied metal AM technologies in both industrial practice and academic research. The use of mAM is increasing; however, is it truly more sustainable than CM? Motivated by this question, a systematic literature review (SLR) was conducted to compare the sustainability impacts of mAM and CM across the three dimensions of sustainability: environmental, economic, and social. The evidence shows mixed sustainability outcomes, which are synthesized later in the conclusions. The sustainability comparison is influenced by factors like part redesign with topological optimization (TO), the material and energy mix used, geometric complexity, production volume per batch, and the boundaries adopted. Economic viability remains critical; companies are unlikely to adopt mAM if it proves more expensive than CM as this could threaten its competitiveness. Social impacts are the least studied dimension, and it is difficult to anticipate the changes that might occur because of such a transition.

1. Introduction

For several decades, the scientific community has warned about the increasing concentration of carbon dioxide (CO₂) emissions in the atmosphere, as shown in Figure 1, which are responsible for climate change and numerous other harmful consequences. The excessive consumption of natural resources poses an additional challenge, as these resources are finite within the Earth’s natural limits.

As illustrated in Figure 1, global CO₂ emissions continue to rise. Among the total global CO₂ emissions, the manufacturing industry accounts for approximately 25%, as shown in Figure 2. This ongoing trend and its associated dangers underscore the urgent need to develop effective solutions.

In recent decades, the global community has increasingly recognized that combating climate change must be a priority due to its wide-ranging repercussions across multiple sectors.

The concept of sustainability emerged in academic discourse with Edward Freeman’s Strategic Management: A Stakeholder Approach (1984) [3]. Subsequently, the Brundtland Commission [4], in a report commissioned by the United Nations in 1987 called “Our common future”, provided the first widely accepted definition of sustainability: “Sustainable development is development that meets the needs of the present without compromising the ability of future generations to meet their own needs.”

Later, John Elkington’s [5] Cannibals with Forks: The Triple Bottom Line of 21st Century Business (1999) introduced the concept of the triple bottom line (TBL), which emphasizes the three Ps: profit, people, and planet. Elkington argued that businesses should not focus solely on financial profit but should also account for their social and environmental impacts. Consequently, the modern understanding of sustainability encompasses three dimensions: environmental (planet), economic (profit), and social (people).

Parallel to the growing concern over CO₂ emissions and the emergence of sustainability as a guiding principle, significant technological innovations occurred during the 1980s. Hideo Kodama pioneered a 3D prototyping machine in the early 1980s, followed by Charles Hull’s invention of stereolithography (SLA) in 1984. In 1991, three new AM technologies were commercialized, including Scott Crump’s fused deposition modelling (FDM) and laminated object manufacturing (LOM). The 1990s were a decade of rapid innovation in AM, with new technologies consistently emerging from research hubs in the United States, Japan, and Europe. Notable milestones from this era include the Fraunhofer Institute for Laser Technology’s (ILT) 1995 patent for selective laser melting (SLM) [6], and the 1997 development of laser engineering net shaping (LENS) [7], a directed energy deposition (DED) technology also known as fused metal deposition (FMD) by the Sandia National Laboratory and Pratt and Whitney (United Technologies Corporation) [8].

According to ISO/ASTM 52900:2021 [9], AM is defined as: “A process of joining materials to make objects from 3D model data, usually layer upon layer, as opposed to subtractive manufacturing methodologies.”

AM technologies share four primary characteristics: they construct objects layer by layer, are based on digital 3D model data, add material only where needed, and can employ a variety of materials, including metals, polymers, and ceramics. When metals are used, the process is referred to as metal additive manufacturing (mAM).

Metal AM offers considerable environmental advantages, including material and waste reduction, lower energy consumption, and the potential for redesigning components to decrease weight and, consequently, fuel usage. These advantages have led many to speculate that mAM could eventually replace conventional manufacturing (CM) processes, which typically rely on mass forming or subtractive methods.

The ASTM and ISO have classified AM technologies into seven process categories in the ISO/ASTM 52900 standard [9]. Among these, the most widely used metal AM processes are defined as follows:

Powder bed fusion (PBF): An AM process in which thermal energy selectively fuses regions of a powder bed.

Directed energy deposition (DED): An AM process in which focused thermal energy is used to fuse materials by melting them as they are being deposited.

Binder jetting (BJT): An AM process in which a liquid bonding agent is selectively deposited to join powder materials.

Metal AM is frequently compared with a wide range of CM processes in the studies included in this review. As observed in the environmental and economic comparison tables (Tables S1–S11), the most common CM technologies used as benchmarks include CNC machining and milling, turning, various casting processes (such as sand casting, die casting, investment casting, and gravity casting), forging, hot forging, rolling, hot rolling, hobbing, MIM (metal injection molding), and electrical discharge machining (EDM). This diversity reflects the broad industrial context in which mAM is evaluated and highlights that sustainability outcomes depend not only on the mAM technology selected but also on the specific CM route against which it is compared.

A review of several studies concerning the sustainability of AM suggests that mAM has significant potential to contribute to CO₂ emission reduction by substituting CM processes. This observation motivates the initial research question (RQ) of the present study: Is mAM truly a more sustainable technology than conventional manufacturing and is there empirical evidence to support this claim? This review finds that the comparative sustainability of mAM versus CM is not absolute but is highly context-dependent, critically affected by variables such as part redesign with topological optimization (TO), production volume, material selected, energy consumption and sources, and boundaries selected. The findings indicate that while mAM can provide environmental advantages, its adoption primarily depends on achieving economic sustainability, with the social dimension of its impact remaining the least explored.

2. Materials and Methods

A systematic literature review (SLR) was adopted due to its structured and transparent process for identifying and synthesizing evidence, as recommended by Snyder [10], Seuring and Müller [11], and Tranfield et al. [12].

2.1. Exploratory Study

The literature search was initially conducted using Google Scholar, employing keywords and topics relevant to the RQ. This search produced 30 papers, many of which were highly cited. Most of the articles were published in Q1 journals according to the SCImago Journal Rank. A significant number of these studies highlighted the potential of mAM to be more sustainable than CM.

2.2. Review Plan and Protocol

The subsequent step in the review process involved developing a research protocol, which included the following elements:

• Objective: To determine whether mAM technologies are more sustainable than conventional manufacturing (CM) methods.
• Definition of conceptual boundaries: The scope of this investigation was limited to metal manufacturing across several industrial sectors, including general industry, automotive, aerospace, space, and dentistry. The study sought to derive its conclusions from empirical data; that is, information obtained through experimentation and measurement.
• Search strategy: The literature search was conducted using two academic databases: Scopus and the Web of Science (WoS). The review focused primarily on papers published in peer-reviewed journals, most of which are ranked in Q1 according to the Journal Citation Reports (JCR) or the SCImago Journal Rank (SJR).
• Inclusion criteria: Peer-reviewed articles and conference papers, published between 2014 and July 2025, written in English, focused on metal manufacturing sectors (general industry, automotive, aerospace, space, dentistry), and containing empirical data.
• Exclusion criteria: Books and book chapters, papers on plastics and construction industries. Books and book chapters were excluded, not because they lack value, but because the objective of a scientific article is to present the latest and most rigorous empirical evidence, which is predominantly found in peer-reviewed journals and conference papers and therefore considered the highest standard of source material.

2.3. Question Formulation

To formulate the RQ systematically, the PICOS framework—population, intervention, comparison, outcome, and study design—was employed. This format is among the most widely used for developing RQ in SLRs and facilitates a more effective database search strategy.

The initial exploratory question that motivated this investigation was refined, with PICOS, into the following formal RQ:

“According to evidence from experimental studies, do mAM technologies demonstrate greater sustainability than CM technologies, considering economic, environmental, and social indicators?”

2.4. Locating Studies

To identify relevant studies, keywords derived from the exploratory search were selected and combined using Boolean operators (OR, AND, NOT). These keywords were organized into three thematic groups: sustainability, mAM, and CM.

This three-group structure, Figure 3, was intentionally designed to align with the PICOS framework defined in Section 2.3 and to rigorously capture the key comparative components identified during the exploratory study (Section 2.1), comparing the intervention (I) of mAM with the analysis (C) of CM based on sustainability outcomes (O) across the metal industry sector population (P) in empirical studies (S).

Before executing the definitive searches, pilot tests were conducted. A total of 27 trials were performed to optimize the search string.

The keywords grouped were as follows:

• Group A (sustainability): LCA, LCC, life cycle assessment, life cycle costing, social impact indicators.
• Group B (mAM): rapid prototyping, metal additive manufacturing, metal 3D printing.
• Group C (CM): conventional manufacturing, subtractive manufacturing, traditional manufacturing.

2.5. Search Query for Reproducibility

• The search was conducted on the Scopus and Web of Science (WoS) databases. The final, search query, utilizing Boolean operators, was performed in three steps to make sure it was accurate and captured all the necessary information from the databases.
• 1. Step: Initial search (Group A):
  The search tool in each database was used to extract all articles discussing the three dimensions of sustainability (“LCA” OR “LCC” OR “life cycle assessment” OR “life cycle costing” OR “social impact indicators”).
• 2. Step: Next refine (Group B):
  Then, the database filter tool was used to reduce the resulting articles, keeping only those that also matched the other two groups (AND “rapid prototyping” OR “metal additive manufacturing” OR “metal 3D printing”).
• 3. Step: Final refine (Group C):
  AND (“conventional manufacturing” OR “subtractive manufacturing” OR “traditional manufacturing”).

The search produced 152 papers in the Web of Science and 185 papers in Scopus, resulting in a total of 337 studies.

2.6. Selection and Evaluation

The following steps were then applied to refine the dataset:

• Duplicate removal: 38 duplicate records were eliminated.
• Application of the predefined exclusion criteria resulted in the removal of 51 papers.
• Identification of false positives: Abstracts were reviewed to exclude papers not relevant to the research focus, resulting in the elimination of 121 papers.
• Identification of doubtful articles: Introductions and conclusions were read to identify papers of uncertain relevance; 33 papers were excluded.

After these steps, 94 articles remained for detailed review. To ensure completeness, false negatives—relevant papers not captured by the initial search—were identified through backward snowballing; i.e., by reviewing the bibliographic references of the 94 selected articles. Papers cited by more than 20% of authors were considered for inclusion, resulting in 12 additional papers, one of which overlapped with the existing list and was therefore removed. The total number of papers selected for full reading was 105.

Following a comprehensive analysis, an additional 20 papers from the initial 105 were excluded, leaving 85. The process of backward snowballing and the inclusion of new key studies added a net total of 8 relevant articles, resulting in a final database used for analysis of 93 articles.

The overall procedure for study selection and refinement followed the PRISMA 2020 [13] flow diagram for new systematic reviews, as summarized in the following chart (Figure 4).

While this paper focuses on empirical evidence comparing mAM and CM, the review also identified several studies that proposed models and frameworks for the comparison.

Ahmad et al. [14] and Hettesheimer et al. [15] propose different models to quantify the energy demand of the process. On the other hand, Ott et al. [16] outlines a conceptual model for a comprehensive activity-based cost assessment. Ribeiro et al. [17] describes a new conceptual framework for life cycle sustainability assessment (LCSA) of products made by AM.

The complete dataset of the studies included in this review, detailing the case-by-case analysis of environmental and economic comparisons, is available in the Supplementary Materials (Tables S1–S11) without restriction.

3. Results

Of the 89 papers identified that made sustainability comparisons, 22 were literature reviews [18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38], and 4 studies propose distinct analytical models for decision support and to quantify energy demand [14,15,16,17]. This left 59 articles that provided empirical evidence comparing the sustainability of mAM and CM, which were selected for detailed analysis.

Of the 59 papers comparing sustainability with empirical evidence, 68% made the comparison in the environmental dimension only, 1.5% in the economical, zero in the social, then 29% in the environmental and economical, and 1.5% in the social and environmental pillars, as shown in Figure 5.

3.1. Environmental Sustainability Results

Tables S1–S6 in the Supplementary Materials present the results of the comparative environmental impact analysis between mAM and CM. The studies analyzed are all based on empirical evidence reported in the literature.

Tables S1–S4 compile case studies that demonstrate an environmental advantage for mAM, detailing the results in four different industrial sectors. Table S1 summarizes findings for the automotive, aeronautical, and naval industries; Table S2 focuses on applications within medical implants; Table S3 covers metal and mold parts; and Table S4 addresses studies concerning industrial and chemical parts industries.

Table S5 presents studies reporting similar environmental performance of mAM and CM.

Table S6 compiles case studies in which CM demonstrated a clear environmental advantage over mAM.

Based on the interpretation of Tables S1–S6, among the 84 cases presenting environmental empirical evidence results, approximately 54% involved PBF and hybrid PBF technologies, while 29% involved DED technologies, including some hybrid processes with DED. Binder jetting(BJ) accounted for only 7% of the cases.

In the environmental comparison between mAM and CM technologies, in approximately 93% of the 84 cases analyzed mAM technologies exhibited lower environmental impacts than CM technologies (Tables S1–S4). In 2% [39,40], both technologies demonstrated comparable impacts (Table S5), while in 5% of the cases [41,42,43,44] mAM showed higher impacts (Table S6), all involving DED technologies.

Table 1. Conditions under which mAM showed lower environmental impact than CM for the most commonly used mAM technologies.

Condition for mAM with Less Environmental Impact	Cases (%)	PBF and PBF Hybrid	DED and DED Hybrid	BJ	MEX and MEX Hybrid
Part Redesign with Topology Optimization	29%	22%	5%	1%	1%
No Specific Conditions Required	23%	8%	10%	1%	4%
Limited Production Batch Size	15%	8%	5%	1%	1%
High Material Removal Required in CM	15%	10%	5%	0%	0%
Use of Renewable/Low-Carbon Energy	14%	10%	1%	0%	3%
Complex Part Geometry	13%	10%	3%	0%	0%
Use of Recycled Material	8%	8%	0%	0%	0%
Extended Part Life/Use Phase Benefits	5%	5%	0%	0%	0%
Other Conditions	30%	17%	10%	3%	0%

Note: The total exceeds 100% due to cases that met several conditional criteria. However, in six cases mAM had more or similar environmental impact (Tables S5 and S6), and in five of these cases it was without any condition. Figure 6 highlights the fluctuating, yet persistent, use of TO.

synthesizes the conditions under which mAM demonstrated lower environmental impacts than CM across the 84 empirical cases reviewed. The most common conditions included part redesign through TO (29%), limited batch sizes (15%), high material removal rates (15%), and the use of renewable or low-carbon energy sources (14%), often in combination. Only about one-fifth (23%) of the cases reported lower impacts without any condition. The use of complex geometries (13%), recycled materials (6%), and extended part lifetimes (4%) also contributed to improved environmental performance. In most cases, these conditions occurred in combination with other factors.

The annual frequency of studies reporting the use of topology optimization (TO) as a condition for mAM’s environmental superiority demonstrated a highly volatile and cyclical pattern between 2014 and 2025, as shown in Figure 6. The data shows a high-amplitude cycle of roughly every two years, with peaks reaching 4 followed by significant lows (0). A linear regression analysis indicated a small positive increase (slope = 0.1329). However, this fitted linear model was not statistically significant (p = 0.334) and explained little of the variability (R² = 0.0935). This indicates that the observed upward trend is unreliable, and the overall variation is dominated by the short-term fluctuations. As per the methodology used for the environmental comparison, 91% used LCA or LCA-based methodologies. A growing trend was observed, as shown in Figure 7.

The analysis of the use of life cycle assessment (LCA) methodology shows a growing trend over the period (Figure 7). The linear trend analysis indicates a positive slope of 0.4680. The model shows a correlation (r) of 0.5966 and an R² value of 0.3559. The p-value is 5.27 × 10⁻². This growing trend is consistent with the fact that 91% of the environmental comparisons used LCA or LCA-based methodologies. Concerning the system boundaries used in the comparison,

Table 2. Scopes employed for the environmental comparisons.

Boundary Scope	Total Cases (%)	Where mAM < CM	Cases (%)
Cradle to Gate	45%	41%	45%
Cradle to Grave	40%	39%	40%
Cradle to Gate	8%	8%	8%
Gate to Use	5%	4%	5%
Others	2%	2%	2%

summarizes how different system boundaries affected the comparative results, indicating the cases in which mAM had lower environmental impacts than CM. The data show that cradle-to-gate (45%) and cradle-to-grave (40%) were the dominant boundary scopes, and that the selected boundary played a significant role in determining whether mAM or CM performed better. In this review, the term cradle-to-gate refers to the system boundary extending from the extraction of raw materials through the manufacturing stage up to the factory gate, whereas cradle-to-grave covers the full product life cycle from the extraction of raw materials, manufacturing, use phase, and finally disposal.

The trend in the most used system boundaries is shown in Figure 8, namely cradle-to-gate and cradle-to-grave.

Regarding the methods used for environmental impact assessment, the ReCiPe method was applied in approximately 31% of the cases, followed by the cumulative energy demand (CED) method in 19%, and IMPACT 2002+ in about 6% of the cases. ReCiPe is a life cycle impact assessment (LCIA) method that harmonizes earlier approaches (CML, Eco-indicator 99, and IMPACT 2002+) to convert emissions and resource use into environmental impact indicators.

The comparison studies present analysis based on case studies drawn from a range of industrial sectors (

Table 3. Distribution of environmental case studies by application area, environmental outcome (mAM vs. CM), and system boundaries employed.

Case Studies	Case Study Application Area	Where mAM < CM	Where mAM ≥ CM	Cradle to Gate	Cradle to Grave	Gate to Gate	Others
26%	Industrial parts	25.00	1.19	19.05%	5.95%	3.57%	0.00%
20%	Metal parts	19.05	1.19	14.29%	1.19%	4.76%	0.00%
18%	Automotive parts	16.67	1.19	2.38%	14.29%	1.19%	0.00%
18%	Aeronautical industry parts	16.67	1.19	3.57%	10.71%	3.57%	0.00%
7%	Mold parts	7.14	0.00	1.19%	3.57%	0.00%	2.38%
5%	Medical parts	4.76	0.00	0.00%	4.76%	0.00%	0.00%
4%	Chemical industry part	2.38	1.19	3.57%	0.00%	0.00%	0.00%
2%	Others	1.19	1.19	2.38%	0.00%	0.00%	0.00%

). The selected cases encompass diverse application areas, including industrial components, metal parts, automotive parts, aeronautical industry components, mold parts, medical devices, and the chemical industry. As shown in Table 3, the analysis of these empirical cases by application area details the methodological scope employed (boundaries) and the environmental result obtained (with mAM < CM or mAM ≥ CM) for every sector, highlighting the context-dependent nature of the findings.

3.2. Economical Sustainability Results

Tables S4 and S5 in the Supplementary Materials present the findings from the economic impact comparisons. Only papers with empirical evidence were included in the cost comparison.

Table S7 compiles case studies from the aeronautical, automotive, and naval industries demonstrating an economic advantage of mAM over CM.

Table S8 presents case studies from the metal and mold industries in which mAM showed lower economic costs compared to CM.

Table S9 summarizes case studies in the industrial and chemical parts sectors that also indicated an economic advantage of mAM relative to CM.

Table S10 includes case studies focused on metal parts where CM demonstrated lower economic costs than mAM.

Table S11 outlines industrial and chemical part case studies in which CM was more cost-effective than mAM.

Seventeen papers evaluated the economic comparison between mAM and CM. Among the 17 papers presenting economic comparisons, a total of 27 case studies were analyzed. Of these, 44% involved PBF and hybrid PBF technologies, 44% involved DED and hybrid DED technologies, 17% involved hybrid technologies, while binder jetting and material extrusion each accounted for 4% of the cases.

In the economic comparison of mAM and CM technologies, 63% of the 27 analyzed cases indicated that mAM is less expensive than CM, while the remaining cases showed higher costs for mAM.

When mAM was considered economically cheaper than CM, only 6% of the cases reported this advantage without any specific conditions. In 47% of the cases, the cost advantage depended on topological weight reduction, typically in combination with other conditions. Notably, 35% of the cases identified limited batch sizes, again in conjunction with additional conditions. Finally, in 18% of cases, the cost advantage was observed solely for large batch numbers. These findings are summarized in

Table 4. Conditions under which mAM is economically cheaper than CM for the main mAM technologies.

Condition for mAM Being Cheaper than CM	Cases (%)	PBF and PBF Hybrid	DED
No conditions required	6%	0%	6%
Part redesign with topology optimization (TO)	47%	12%	35%
Limited batch number	35%	6%	29%
Batches with more than a given number of parts	18%	12%	6%
Other	41%	24%	18%

.

Table 4 shows that mAM became economically competitive mainly when redesign through TO was applied (47% of cases) or when batch sizes remained small (35%), confirming that mAM becomes cost-effective only under certain conditions and is not universally cheaper than CM.

When mAM was more expensive than CM, 80% of the cases were attributed to the high cost of machinery. In 40% of the cases, the cost of metal powders was the main factor, with some cases affected by both conditions simultaneously.

Regarding the methodology used for economic comparisons, 52% of the cases employed life cycle costing (LCC) or LCC-based methods, while the remaining cases applied other approaches.

Among the case studies, 44% involved industrial parts, 19% metal parts, 15% aeronautical components, 11% parts for molds, 8% parts from the chemical industry, and 3% naval components.

3.3. Social Sustainability Results

The social sustainability dimension remains an insufficiently researched area in the comparison of these manufacturing technologies. To date, one of the few studies providing empirical evidence is by Cappucci et al. [45], who performed a social comparison based on a case study of Ti6Al4V alloy femoral stems. The study compared LPBF with CM (forging + milling + machining). The analysis demonstrated that LPBF was more socially impactful in product performance and in industrial product function utility. The authors state that mAM has more social impact than CM because mAM is able to create a more complex and biocompatible geometry of the stem, which is impossible for CM technology to achieve. Worker health and safety emerged as the most significant social concern in the literature, due to the hazards associated with inhaling fine powders and exposure to chemical emissions such as VOCs during processing [18,19,29,32,35,37].

Many papers conducting literature reviews also found that the social pillar was the most neglected and understudied area of the AM sustainability comparisons [20,26,30,32]. Also, several reviews noted the scarcity of research in this dimension, highlighting the lack of standardized methods to assess the social impacts [21,26,28,29,37].

4. Discussion

4.1. Environmental Sustainability Discussion

The finding that mAM exhibits a lower environmental impact is particularly noteworthy given that PBF was the most frequently analyzed technology in the reviewed studies, with roughly twice as many assessments as DED. PBF is widely recognized as the most energy-intensive mAM route owing to its comparatively slow build rates, fine layer resolutions, and extended exposure times [39,46,47,48]. This observation is consistent with previous literature reviews that highlight the substantial energy demands associated with mAM processes [20,22,26,34,35,38] as well as the significant environmental burden of feedstock production [20,37]. Nevertheless, the higher process energy requirements of PBF can be diminished by the benefits enabled through TO. In particular, TO can produce substantial reductions in use-phase impacts—especially within the aeronautical and automotive sectors—thereby often resulting in a net environmental advantage over CM.

Only a small fraction of studies reported unconditional environmental benefits, highlighting the importance of situational conditions. This aligns with the patterns summarized in Table 1, where redesign, production volume, geometric complexity, and energy mix consistently determined whether mAM outperformed CM. The frequent use of TO shows that design optimization has been a key contributor across the years but we suggest interpreting this trend cautiously. The relevance of TO is in enabling lightweight design, resulting in environmental benefits observed in mAM. Although TO appears frequently in recent work, the variability of weight designs is particularly notable in sectors such as automotive and aeronautics, where use-phase efficiency gains are significant. These findings are consistent with previous literature reviews highlighting AM’s advantages in reducing material waste and enabling lightweight designs that lower energy consumption during product use [21,22,23,24,28,35].

However, this review revealed an apparent discrepancy: despite the well-established benefits of TO, its documented use in the automotive and aeronautical sectors appears irregular, as illustrated in Figure 9. This observation must be interpreted with caution, since the applied search filter (requiring LCA or LCC terms) may exclude technical studies in which TO is applied but not accompanied by a full sustainability assessment, which could therefore lead to an incomplete representation of its actual industrial use.

Figure 9 shows substantial year-to-year volatility, with pronounced peaks (e.g., 2016, 2021, 2024) and intermittent periods of low publication activity. Although the fitted linear trend line exhibits a slight negative slope (m = −0.0606), the very low coefficient of determination (R² = 0.0165) and high p-value (0.724) indicate that this trend is not statistically significant. Thus, the data do not support a meaningful decline in TO applications; rather, they highlight the irregular and fragmented reporting of such case studies over the 2016–2024 period. It is essential to note that the high volatility and statistical insignificance of this trend are likely a result of the systematic literature review’s search filter, which required the presence of sustainability terms like LCA or LCC. This methodological constraint likely excluded technical case studies where TO was applied but a full sustainability assessment was not conducted, leading to an incomplete representation of its actual industrial use. The full implications of this search constraint are detailed in the Limitations section (Section 4.4).

Although the condition of reusing material is not one of the most common in this paper, most authors recognize mAM’s strong potential for material reuse and circular economy benefits through powder recycling, repair, or remanufacturing. mAM’s near-net-shape production and design flexibility reduce waste and material use, supporting sustainability and resource efficiency. A strong potential for circular economy integration is observed. In the literature review, several papers also underline the potential of mAM to reinforce the recycling strategies and the circular economy models ([19,28]).

In contrast, many studies note limits on powder reuse due to degradation, oxidation, or contamination after several cycles, although some papers report successful multiple reuses, emphasizing quality monitoring and process control to ensure material integrity during recycling. The limited reusability of materials due to degradations is also noted in the literature review by Arrizubieta et al. [18] and Moura et al. [29].

LCA is the most used method to assess the environmental impact in the studies, and its use is also growing, as shown in Figure 7, although some papers in the literature review consider a major barrier to be the lack of standardized life cycle assessment methodologies [26,34] and reliable inventory data [27,32]. This finding diverges from the claims in other reviews that LCA methodologies lack standardization.

Cradle-to-cate and cradle-to-grave approaches in the reviewed studies introduce variability in the reported outcomes, as the inclusion or exclusion of the use phase can alter the relative performance of mAM and CM. The revised Table 3 clearly reinforces this challenge by quantifying the environmental outcome per sector. It shows that sectors dealing with highly complex parts, such as medical parts and mold parts, exhibit a greater concentration of environmental superiority for mAM, with 4.76% and 7.14% of cases, respectively, showing mAM < CM and zero cases demonstrating mAM> or =CM. This outcome is consistent with AM’s advantage in realizing complex geometries, which CM cannot achieve. Conversely, the industrial and metal parts sectors, which often involve fewer complex geometries, showed a higher percentage of cases where mAM was only marginally superior or comparable/worse to CM (1.19% of cases show mAM ≥ CM for both industrial and metal parts). The sectoral distribution shown in Table 3 also highlights that industrial, metal, automotive, and aeronautical cases predominantly rely on these two boundary types (cradle-to-gate and cradle-to-grave). This relationship helps explain why environmental results vary substantially across sectors and highlights the need for consistent boundary selection to improve comparability across studies (Rev. 2).

The choice of system boundary should reflect the actual life cycle of the part, consistent with previous literature showing that methodological decisions strongly influence comparative results [25,29,36,37]. Cradle-to-grave boundaries are particularly important when AM-enabled design modifications affect the use phase—such as lightweighting achieved through TO—because these benefits are not captured in cradle-to-gate assessments.

4.2. Economical Sustainability Discussion

A disproportionate amount of research has focused on environmental impact (analyzed in 98% of studies) while far less attention is paid to economic comparison (31% of the studies). This reveal a critical fact and suggests that the academic community is not prioritizing the most important key to mAM adoption which, as this review confirms, is economic viability.

This economic analysis should be the first step taken in the adoption of a mAM technology vs the traditional CM. No company will adopt a modern technology that has less environmental impact or that has more positive social impact if it is more expensive. The reason has to do with the loss of competitiveness of the company in highly competitive markets. Several authors agree in their conclusions: “The decision for choosing a manufacturing technology for a specific product is primarily based on cost in industrial practise.” (Kamps et al. [49]); “The scarcity of available literature on the cost-efficiency and lacking reports from industrial scientific sectors currently limit AM industrial full implementation” (Nyamekye et al. [50]).

The conditions for mAM to be less expensive than CM vary with the cases studied and the conditions adopted. The results show that the redesign of parts with TO for reducing weight, the production of small batches or the combination of both are the most used conditions to achieve cost competitiveness. Across sectors, 63% of the cases that combined mAM with TO and adopt cradle-to-grave boundaries reported environmental and economic benefits in aeronautical and automotive applications. By contrast, CM outperformed mAM in 81% of general industry cases where no redesign or extended boundaries were applied. These results indicate that mAM’s advantages are context-dependent and rely on redesign, batch size, and the selected system boundary. Table 4 makes this conditionality explicit, showing that cost advantages arise mainly when redesign or low-volume production is feasible. This concentration of benefits in sectors such as aeronautics and automotive reflects the high value of weight reduction and the practicality of redesign in these industries.

Metal AM becomes cost-competitive at low batch sizes because it does not require tooling, which represents a major cost in CM. Topology optimization (TO), occurring in 47% of the cases, further improves cost performance by reducing material use and helping to compensate for the higher base cost of mAM. These findings show that mAM is most cost-effective in low-volume production and when substantial material savings are achieved through TO. However, economies of scale continue to make CM cheaper for mass production, which explains why most studies find mAM to be more expensive when no specific conditions are applied. Nevertheless, none of the reviewed cases mention that a condition for mAM to be more expensive than CM is for large-scale production. As stated by Hegab et al. [23], “AM not only cannot compete with traditional manufacturing for mass production, but it is also unsuitable for larger batch production systems.” This finding agrees with the literature review, which shows that CM is typically more economical for high-volume production of simple parts due to economies of scale [26,29,30,31]. This lack of competitiveness for large production quantities represents a lost opportunity in the automotive industry.

For mAM to be more expensive, the major cause is the cost of machinery. The metal powder production costs is also another of the main reasons. This is because producing metal powder involves an additional cost, since it employs the same ingots used by CM having then an additional process. This is consistent with the results in the literature reviews, which find that the path to widespread adoption is obstructed by significant economic barriers, chief among them being the high initial capital investment in machinery and the steep cost of specialized materials and energy [20,22,35]. While PBF and DED are technically better for complex parts, these high costs for machinery and expensive powde, made them more expensive in most general comparisons with other technologies. Nevertheless, there is a mAM technology with a potential cost advantage, which is binder jetting (BJ). This involves less complex and expensive machinery than laser-based PBF and allows for higher powder reuse. These two points can offer a path to cost reduction for certain geometries and materials.

In contrast to LCA, which remains the most widely used method, LCC was applied in only half as many studies. This limited adoption reflects the lack of comprehensive and standardized economic data, posing a barrier to reliable cost comparison between mAM and CM. This agrees with the results obtained in the literature review findings from some authors that emphasize that high machine and material costs still hinder broader competitiveness, a challenge magnified by the lack of comprehensive and standardized data needed for reliable economic modelling [37,38,51].

Due to the importance of mAM being cheaper for its adoption in industry, it is necessary to work on cost reduction, which depends on deposition rates, build orientation, manufacturing times, printing multiple components at the same time, using TO to reduce material usage, energy efficiency, and material recyclability. Some studies, however, indicated limited additional savings with current technologies and process maturity levels.

Another way to reduce mAM costs is to use a distributed AM model where producing locally on-demand eliminates transportation costs and reduces lead times compared to centralized CM, making it more economical overall. This aligns with the results of the literature review of some authors, who found that this alternative can cut inventory costs, and enables innovative circular economy business models centered on the repair and remanufacturing of high-value components [23,29] and emphasizes AM’s potential to enable new business models and transform value chains [20,21,30].

4.3. Social Sustainability Discussion

Regarding social impact, there is a consensus that it is the least-studied aspect of sustainability, making it difficult to obtain clear results. This matches the findings in other literature reviews, with the social pillar consistently identified as the most neglected and under-researched area of AM sustainability [20,26,30,32].

This scarcity of research is complicated by a lack of standardized assessment methods. Several literature reviews explicitly note the scarcity of research and standardized social life cycle assessment (S-LCA) frameworks to assess social impact as a critical barrier to a comprehensive understanding of these complex impacts [21,26,28,29,37]. To move toward a more standardized methodology, future comparative studies can take advantage of existing frameworks, such as the UNEP (United Nations Environment Programme)/SETAC (Society of Environmental Toxicology and Chemistry) guidelines for S-LCA [52], to structure their assessment and increase comparability.

Therefore, the analysis of social impacts should be structured centered on stakeholder group data, based on the currently available empirical information. One of the stakeholder groups is workers. The highest concern highlighted by several authors is worker health and safety and its management. This results from the risks of inhaling hazardous fine powders and exposure to chemical emissions like VOCs during the manufacturing process [18,19,29,32,35,37].

The other stakeholders are consumers and society in general. The positive social impacts identified include product utility. Pusateri et al. [32] demonstrate with their medical device comparison that mAM has more social impact than CM because the system studied was impossible to develop using CM technology. This demonstrated that LPBF has a greater social impact than CM because it can create a more complex and biocompatible geometry. This positive contribution is often highlighted in the literature for healthcare through customized medical devices [24,30,33,35,38].

In summary, the social dimension is constrained by the difficulty of obtaining empirical data, because many evaluations rely partly on qualitative assessments. Research must continue to advance in this area, particularly concerning the health risks to workers and the tangible positive effects on consumers and society.

4.4. Limitations

The search strategy, which required the presence of sustainability terms like LCA, LCC, etc., to identify comparative studies, may have inadvertently filtered more recent technical implementation papers that utilize TO in these sectors but do not conduct a full LCA/LCC comparison. This search constraint, in particular, affects the reliability of observed adoption trends, such as the apparent decline or volatility in TO use shown in Figure 9. A further limitation is that published case studies frequently analyze scenarios where redesign or small-batch production is already suitable for mAM, which may overrepresent contexts favorable to the technology.

5. Conclusions

Interest in comparing the sustainability of mAM and CM technologies is growing. The evidence synthesized in Table 1, Table 2, Table 3 and Table 4 indicates that sustainability outcomes vary according to redesign potential, production scale, sectoral characteristics, and methodological choices inherent to the technologies themselves. The review highlights four recurring insights: the high energy demand of mAM, the predominant use of PBF and DED, the relevance of design optimization, and the economic importance of production scale.

Environmental advantages emerge mainly when mAM enables lightweight benefits (especially in aerospace/automotive) for complex designs that reduce use-phase impact.

Economic viability remains the primary barrier to widespread adoption. mAM is economically advantageous over CM primarily for low-volume production batches and for TO-optimized parts, while CM remains more cost-efficient for large-volume fabrication due to economies of scale.

Social impacts are the least studied dimension, due to the lack of standardized methods. Key concerns center on worker health and safety due to fine powder and VOC exposure, while positive contributions are most evident in customized medical applications.

6. Recommendations

Based on the findings of this review, the following recommendations are essential to support future research and guide the sustainable adoption of metal AM:

• Prioritize economic cost reduction: Research must focus on improving cost-efficiency. This requires reducing the high initial capital investment in mAM machinery and the steep cost of metal powders. Costs can be lowered through improved deposition rates, optimized build orientation, increased energy efficiency, and greater material recyclability.
• Integrate LCC and LCA: Future studies must integrate life cycle costing (LCC) alongside life cycle assessment (LCA). Since economic viability is the main barrier to industrial adoption, demonstrating cost-effectiveness is essential before environmental or social benefits can support implementation. Broader use of LCC is needed.
• Adopt cradle-to-grave standardization: A cradle-to-grave system boundary should be adopted whenever possible, as it is the only way to capture the critical long-term benefits resulting from weight reduction and use-phase efficiency.
• Advance social impact quantification: Future research must establish frameworks like the UNEP/SETAC guidelines to develop clearer, quantitative S-LCA indicators. The immediate priority is to quantify risks related to worker health and safety (powder and VOC exposure).
• Define clear targets for sustainable mAM adoption: Industry and researchers should collaborate to define measurable performance standards, such as minimum weight-saving targets or maximum viable batch sizes, to clarify the conditions under which mAM is guaranteed to surpass CM.