AI-Driven Innovations in 3D Printing: Optimization, Automation, and Intelligent Control

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

1、No explicit search strategy (databases, time window, keywords), inclusion/exclusion criteria, or study quality assessment.

2、Findings rely on case-by-case narratives; no normalized comparison across datasets, metrics, latency, or data scale.

3、The discussion of DED/WAAM is shallow compared with PBF/FDM; little on multiphysics coupling, bead geometry/height control, dilution, thermal accumulation, residual stress, or MIMO control.

4、Limited analysis of data governance, IP/privacy, model robustness/generalization across machines/materials, calibration/retuning effort, and alignment with quality standards in regulated sectors.

Author Response

Please see the attachment.

 

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The paper presents a timely and comprehensive review of AI-driven innovations in 3D printing, effectively synthesizing recent advancements in optimization, automation, and intelligent control. However, several critical limitations warrant attention.

Firstly, the review lacks a systematic framework for evaluating the maturity and scalability of the discussed AI technologies. While numerous case studies are cited, there is insufficient critical analysis of their real-world applicability, computational costs, or barriers to industrial adoption. For instance, claims about "91.7% mAP accuracy" in defect detection lack discussion of validation methodologies, dataset diversity, or comparative baselines, leaving readers uncertain about practical reliability.

Secondly, the paper overlooks significant ethical and security implications of AI-integrated 3D printing, such as data privacy risks in IoT-enabled systems, algorithmic bias in generative design, and vulnerabilities to cyber-physical attacks in smart factories. These omissions are particularly concerning given the emphasis on healthcare and aerospace applications, where safety-critical consequences demand rigorous risk assessment.

Thirdly, the economic feasibility of AI-driven 3D printing is not fully explored. The review highlights efficiency gains but neglects cost-benefit analyses of implementing AI infrastructure. For resource-limited settings like small-scale manufacturers or developing regions, the prohibitive costs of AI-hardware integration and specialized personnel training could exacerbate technological disparities.

Fourthly, the paper inadequately addresses material-specific AI limitations. While ML/DL applications for metals and polymers are detailed, other materials receive cursory treatment, despite their unique challenges in printability prediction and defect dynamics. The omission of recent work on AI for non-Newtonian fluid extrusion or bioactive ink optimization represents a missed opportunity to showcase cross-material adaptability.

Fifthly, the review’s structure occasionally prioritizes breadth over depth. Sections like "Reinforcement Learning (RL)" and "Computer Vision" list examples but fail to critically examine shared limitations, such as sim-to-real gaps in RL policies or the dependency of vision systems on ideal lighting/angle conditions. Similarly, the industrial impact section catalogs sector-specific applications but does not quantify AI’s ROI in case studies like GE Aviation’s fuel nozzles or Organovo’s bioprinting, weakening the argument for commercial viability. Finally, the conclusion overstates AI’s current capabilities, using deterministic language like "revolutionizing" and "transformative" without acknowledging persistent challenges: data scarcity for rare defects, poor generalizability across printer architectures, and the "black-box" nature of deep learning models that hinders troubleshooting in mission-critical prints. A more balanced perspective, acknowledging both progress and open research questions, would strengthen the manuscript’s scholarly contribution.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The paper presents a review of the integration of artificial intelligence (AI) into 3D printing systems across a wide range of applications. The manuscript covers a relevant topic but requires substantial revision in both content and structure. Many sections lack clarity, are poorly organized, or contain technical and grammatical inconsistencies. I recommend major revisions prior to reconsideration for publication. While the topic is indeed relevant and timely, the manuscript appears to be in an early stage of development. Numerous errors and unclear passages need to be addressed before the paper can be considered for publication.

Major Comments:

Structure of Sections 4 and 5:
I recommend significant revision of Sections 4 and 5. At present, there is little distinction between them. Both sections primarily enumerate previously published works and briefly describe the results, without a clear difference in scope or focus. For instance, Subsection 5.1 Real-Time Process Optimization includes studies involving artificial neural networks (ANNs) and Gaussian models—topics that are also covered in Section 4.1.1 Machine Learning (ML). To improve clarity and organization, I strongly encourage the authors to merge Sections 4 and 5 into a single unified section. This section could be organized by AI category (e.g., ML, ANN, fuzzy logic) or by application domain (e.g., defect detection, process control, topology optimization), but it should avoid repetition and overlapping content.

Other specific Comments:

Line 105:
The authors mention "thermoplastics in material extrusion." It is unclear why thermoplastics are presented in a seemingly negative context. Are the authors suggesting that the use of thermoplastics is a limitation? If so, please clarify whether there are other material families (e.g., thermosets, metals, ceramics) that are considered more favorable for extrusion-based printing. The current phrasing is ambiguous.

Line 118:
The discussion of SLA/DLP technologies references limitations in material variety. This point needs further explanation. Please specify what these limitations are and whether they refer to resin chemistry, mechanical properties, or compatibility with functional additives.

Line 122:
Please clarify the sentence: “After the printing process, the cleaning step is also a major concern (Hassanpour et al., 2024).” What specifically makes the cleaning process problematic—time consumption, cost, health risks, or technical complexity?

Line 141:
The sentence “Material jetting devices are not suitable for large parts” is not entirely accurate. For example, the EOS P 770 offers a large build volume that suffices for many industrial applications. Consider revising this claim or providing a quantitative comparison with other technologies to substantiate the statement.

Section 3:
This section lacks appropriate formatting and cohesion. It should be completely rewritten as a series of well-structured paragraphs discussing the main challenges in the field. While cost is indeed a common issue in any manufacturing technology, the authors should contextualize this by comparing it with traditional manufacturing costs—for instance, the expense of high-speed machining a titanium alloy block. The current presentation is fragmented and superficial.

Lines 216–217:
These lines are not an appropriate conclusion to Section 3. They appear to attempt a summary but fail to convey a clear, focused message. Please revise to conclude the section with a concise summary of key challenges.

Lines 230–232:
There are at least two missing references in this section. Please ensure that all claims and examples are properly cited with recent and relevant sources.

Line 237:
The term “3D micro-printing” is introduced without explanation. Please provide details about the specific technology involved—e.g., two-photon polymerization, micro-SLA—and not just the scale or resolution.

Line 243:
This paragraph addresses the tensile strength of 3D printed products rather than the 3D printing process itself. Consider relocating this discussion to a section on mechanical properties or post-processing analysis.

Line 255 / Figure 9:
Please clarify what type of data is presented in Figure 9. Are the results derived from the authors’ own experiments or from literature sources? This must be explicitly stated.

Section 6: Smart Manufacturing Systems:
This section cannot be adequately reviewed due to multiple missing references, such as:

L846–L868: Error! Reference source not found.

L885: Error in Reference.

L905: Error in Reference.

Please resolve all citation and referencing issues before resubmission.

Final minor Comments and Style Issues

Acronyms:
There are several inconsistencies in the use and formatting of acronyms throughout the paper. Please ensure that each acronym is defined upon first use and used consistently. Examples include:

Material Extrusion (FDM) → Should be Fused Deposition Modeling (FDM)

Artificial Intelligence (AI) → Use lowercase when appearing in running text: artificial intelligence (AI)

machine learning algorithms → Consider using ML algorithms for consistency

Internet of Things → Should be Internet of Things (IoT) upon first mention

General Acronym Use:
A thorough review of the entire manuscript is recommended to ensure all acronyms (e.g., Random Forest, CNN, FDM) are properly introduced and formatted. Consistency in technical terminology is essential for professional and academic clarity.

 

Comments on the Quality of English Language

A thorough review of the entire manuscript is recommended to ensure all acronyms (e.g., Random Forest, CNN, FDM) are properly introduced and formatted. Consistency in technical terminology is essential for professional and academic clarity.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

The authors present a review paper on a very interesting and up to date topic. I will recommend it for publication after some revisions.

Mainly I would like to see the entire paper condensed as I feel its too long with 38 pages and somewhat repetitive. I suggest the use of comparative tables and diagrams between the methods and technologies in multiple sections. Also, emphasize the Industry 4.0/5.0 context in the entire paper. I also suggest the use of the original figures and schemes, especially for AM technologies.

Additional comments:

1. Abstract is descriptive with little critical insight, highlight contribution. State what this review does differently.
2. Wikipedia can’t be used as a source of references.
3. In the Introduction present motivation: Why is AI essential in AM now? Clearly state objectives.
4. Too long, overly descriptive of AM processes. This is already well-known. Expand challenges section to directly link to AI.
5. Fix all “Error! Reference source not found.” entries.
6. Add explicit discussion of unresolved challenges/research gaps that future work must address like: Data scarcity and lack of public datasets, Generalizability across printers/materials, Certification barriers in aerospace/medical AM, Computational costs for real-time deployment, Trust and explainability of AI decisions.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

revision ok

Author Response

We would like to express our gratitude to all the reviewers.

Reviewer 3 Report

Comments and Suggestions for Authors

The authors addressed the reviewer's comments and improved the article. 

Author Response

We would like to express our gratitude to all the reviewers.

Reviewer 4 Report

Comments and Suggestions for Authors

Thank you for the revisions made. I have no further comments.

Author Response

We would like to express our gratitude to all the reviewers.