AI Integration in Fundamental Logistics Components: Advanced Theoretical Framework for Knowledge Process Capabilities and Dynamic Capabilities Hybridization

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

A lot of things need to comprehensively discussed in the paper before it can be recommended for publication. Here are my comments:

1. Where is the structural diagram of the proposed theoretical framework?
2. Present in the methodology the specifics of the research design including how both exploratory factor analysis and confirmatory factor analysis were conducted. What data were used in the EFA, CFA, and other areas in your analysis? How were the testing and validation performed?
3. What cluster analysis method did you use? There is a variety of clustering techniques including partitional, hierarchical, spectral, etc...what were the metrics used to judge the cluster quality?
4. Could you clearly explain the entire flow of the data analysis? It would be better if you could include a flow chart depicting the entire methodology.
5. How is Table 3 formed? In particular, how did you get the values depicted in Table 3? Were these predicted or are they simply descriptive summaries obtained from the data? For example, how do we know that the risk level of the "radical configuration" is "high"? It's not clear how these values were derived. The same can be said with Table 4.
6. In your paper, you claimed you are proposing a theoretical framework, but I  could not find any framework in your paper. There should at least be a diagram that embodies the framework being proposed.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Introduction

The introduction reads more like an unstructured annotated bibliography than a focused argument leading to a clear research gap.

It attempts to cover too many disparate topics, including AI in logistics, Industry 4.0/5.0, knowledge-based view (KBV), dynamic capabilities (DC), sustainability, circular economy, and institutional theory.

This introduction attempts to do too much, confusing the reader and diluting the contribution. I strongly recommend that the author separate the literature review into a dedicated section.

Claiming that “existing literature inadequately addresses…” (lines 101, 124, 231, 236) without citing seminal works or demonstrating how this paper will advance theory is insufficient.

The review cites over 30 references in the introduction alone, yet the integration of these works is superficial.

Conclude the introduction with clear research objectives, questions, and a rationale for the framework.

There is significant repetition of key phrases and ideas, such as “AI systems possess autonomous learning and decision-making capabilities,” “transform traditional knowledge processes,” and “co-evolutionary development.” These phrases appear in various forms multiple times (e.g., lines 97, 104, 234, 268, 276), suggesting a lack of editorial rigor and weakening the impact of the core arguments.

Material and methods

There is an attempt to impress through methodological breadth (e.g., multi-group SEM, CFA, mixed-methods, cluster analysis, thematic coding, longitudinal waves), yet the manuscript fails to explain how these components are logically connected or practically implemented.

The section feels lacks a clear, step-by-step narrative. It is not transparent how each methodological choice supports the research questions or hypotheses. This undermines reproducibility and interpretability.

The authors provide detailed psychometric statistics (e.g., α, AVE, CR) but offer no information on actual items, example questions, or survey instruments.

The authors propose a multi-stage, stratified random sampling of 450 organizations across Brazil, India, China, Mexico, and Eastern Europe. This is a massive and costly endeavor requiring extensive field access, localized partnerships, and cross-cultural coordination.

There is also no description of how such a diverse and complex sample was realistically accessed, raising doubts about the authenticity and feasibility of the data collection process.

The manuscript states that a longitudinal design was used, with “multiple data collection waves.” However, there is no timeline, no details on wave intervals, no panel retention rates, or examples of longitudinal variables tracked.

The reported CFA and SEM model fit indices are acceptable, but the model is not visualized, and theoretical justifications for paths are vague. Hypotheses are introduced with arbitrary coefficients (e.g., β > 0.3) as if predetermined, which raises questions about p-hacking or post-hoc theorizing.

Results

Despite being labeled as a "Results" section, the content does not present any original empirical findings, experimental outcomes, or data analyses. Instead, it consists almost entirely of theoretical generalizations, secondary literature summaries, and speculative assertions.

Table 3 presents a supposedly analytical framework comparing AI implementation pathways, but there is no explanation whatsoever of how the figures were derived—no data source, calculation method, or citation is given.

The section is heavily descriptive and reads more like an extended literature review than a report of findings.

The frequent citations (e.g., [6], [38], [19], etc.) further blur the line between what has been discovered in this study and what is known from prior work

There is no effort to organize the supposed results around a clear analytical framework, set of themes, or research model.

The section jumps between warehousing, transportation, ecosystem orchestration, and innovation without any guiding structure or transitions.

Please provide a concrete example or sample calculation (ideally as an appendix or supplementary file) to illustrate how key performance indicators (e.g., ROI timeline, success rate, scalability index) were derived. This is particularly important for Table 3, where numerical values are presented without any explanation of how they were calculated or sourced.

The current results are heavily narrative and conceptual in nature. Incorporating quantitative data—such as descriptive statistics, comparative performance metrics, simulation results, or case-based figures—would significantly strengthen the empirical foundation. Even a basic quantitative comparison across AI applications (e.g., warehousing vs. transportation) would add clarity and credibility to the discussion.

Discussion

he section fails to offer a substantive, evidence-based discussion of the study’s actual empirical results. While the manuscript claims to use a mixed-methods approach, Section 4 reads almost entirely as a theoretical essay rather than a data-driven interpretation.

The section asserts multiple contributions to organizational theory, digital transformation, and AI adoption, but fails to substantiate these claims with rigorous argumentation or original conceptual development. For example, the supposed extension of dynamic capabilities theory to the ecosystem level is asserted without adequate theoretical scaffolding, citation of relevant boundary-spanning frameworks, or comparative analysis with existing literature.

Virtually every concept discussed—dynamic capabilities, knowledge management, digital infrastructure, platform strategy, regulatory flexibility—has been extensively addressed in existing research, including in emerging market contexts.

I strongly recommend the authors add a dedicated section—either standalone or integrated into the conclusion—that addresses Managerial Implications and Practical Implications.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The authors studied the related issues of AI empowering logistics transformation. But some problems should be taken into account.

1. The abstract is too general. The authors should provide: research background, problem, solutions, experimental results, conclusion.
2.  The introduction is too long. Please use subtitles to address each content.  Like: 1.1 The research background;  1.2 The main problems of the research; 1.3 The solution of the problems; 1.4 The innovation and limitation of the proposed method; etc.
3. The related work must be arranged in another section "2. Related Work" . Since the introduction is too long. In the mean time, the comparison between your proposed work with the previous work must be compared, either in text, or in a comparison Table. 
4. The section "Materials and Methods " also has the same problem. It is too long to read. Use "3.1 The whole research Step", "3.2 Research Material", "3.3 Research Method (proposed work)", etc., to organize the content.    The section 3.1 The whole research Step, the authors should  provide a research flow chart.
5. The section "Results" and "Discussion " should also be arranged by subtitles. Especially, the conclusions should be given in the form of "one-by-one".
6. The comparison experiment should be provided to verify the conclusion: " Findings reveal that hybrid capability architectures, characterized by synergistic integration of knowledge process capabilities and dynamic capabilities, demonstrate superior AI adoption effectiveness compared to traditional capability configurations. "
7. In conclusion section, the following contents should be arranged: (1) The research conclusion; (2) The advantage, innovation and limitation of the work; (3) The future work.  And use the subtitles to arrange.

 

 

 

 

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

While some of my comments were addressed. Three comments were still not addressed. Because of that, I believe another round of revisions would be needed for the paper to be published in the journal:

1) What is the metric used for cluster quality? Did you use Silhouette Coefficient, Davis-Bouldin Index, etc? How did you ensure that the generated clusters were really high quality clusters?

2) It was not explained why the authors opted for a k-means clustering approach. Why not spectral clustering? What certain considerations did you use to arrive at the decision of using k-means clustering? For example, k-means works best when clusters are globular. What evaluations were made to assess this?

3) The presented figure still does not look like a structural diagram of the theoretical framework. It would be best if the paper could show the specific relationships wanted to be tested by the paper. For example, a structural diagram of the factor analysis performed could be shown. 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Introduction

I appreciate the significant improvement made in the revised Introduction. The section now presents a clear and compelling narrative that positions artificial intelligence in logistics (AI-L) as not merely an operational tool but a transformative enabler of adaptive, self-organizing ecosystems. The integration of Industry 4.0 dimensions such as IoT, cyber-physical systems, and cognitive computing is well articulated and effectively contextualizes the disruptive potential of AI-L. Moreover, the introduction of the hybrid capability (HC) framework is a notable advancement, as it demonstrates conceptual depth by synthesizing the Knowledge-Based View (KBV) with Dynamic Capabilities (DC) theory. This is a strong contribution that differentiates the study from prior work.

The articulation of the research objectives and research questions has also improved, providing the reader with a structured pathway to follow the logic of the study. The alignment of prior studies (Table 1) with the present research focus is particularly commendable, as it clarifies the paper’s unique contribution. The layering of the framework (antecedents, mediating processes, moderators, and outcomes) shows a well-thought-out conceptual design that will resonate with both academic and practitioner audiences.

 I would suggest minor revisions for further refinement:

 While the theoretical depth is impressive, some sentences remain very dense (e.g., lines 28–45). Shortening or breaking them into smaller units would improve accessibility for a broader audience, especially practitioners who may benefit from your framework.

The introduction could briefly highlight one or two real-world logistics challenges in emerging markets (e.g., fragmented infrastructure, limited digital readiness) to ground the theoretical narrative and emphasize the practical urgency of AI adoption.

The phrase “hybrid capability frameworks” is a strong and innovative concept. However, a brief comparison with similar constructs in prior AI-logistics or capability literature would further sharpen how this framework stands out as novel.

The flow from the research questions (lines 52–56) to the hypotheses (Figure 1, Table 1) could benefit from one additional bridging sentence, ensuring a smoother transition into the empirical component of the paper.

Literature Review

The revised literature review section demonstrates substantial improvement in clarity, depth, and organization. The authors have successfully expanded the discussion of emerging markets (EMs), knowledge process capabilities (KPC), and dynamic capabilities (DC), linking them to AI-enabled logistics integration with stronger theoretical grounding. I particularly appreciate the way the review now synthesizes multiple theoretical perspectives (capability theory, organizational learning, and institutional theory) and integrates recent empirical evidence. The identification of research gaps is clearer and more compelling, especially regarding knowledge management, capability evolution, and integrated frameworks for AI adoption in EMs. The addition of ethical, social, and institutional considerations also enriches the discussion and broadens the contribution.

For further refinement, I suggest minor revisions: (1) streamline some lengthy paragraphs by grouping related arguments to improve readability, as certain sections are dense with overlapping concepts; (2) ensure consistency in the treatment of hypotheses (H1–H6) by briefly connecting them back to the reviewed literature, making the logical progression from literature to hypothesis development more explicit; and (3) strengthen transitions between subsections (e.g., from AI applications in logistics to institutional challenges) to guide the reader more smoothly. Addressing these minor issues will make the literature review more cohesive and accessible while preserving its comprehensive coverage.

Results

The revised manuscript shows substantial improvement in both theoretical framing and empirical discussion. I particularly appreciate the clearer articulation of knowledge-based view (KBV) and dynamic capability (DC) integration and the stronger contextualization across emerging markets (Table 7). The expansion of methodological details and acknowledgement of limitations also enhance transparency. For example, the inclusion of longitudinal subsample data (n=127) demonstrates a commendable effort to strengthen causal inference.

That said, a few minor revisions would further improve clarity and impact:

 At times, the manuscript alternates between "AI-enabled logistics organizations," "hybrid implementers," and "AI adopters." Standardizing the terminology throughout would improve readability.

While Tables 6 and 7 are informative, captions could be expanded to briefly highlight their interpretive value (e.g., how demographic balance influences findings or how cross-market variance relates to AI adoption).

 The discussion is comprehensive, but the comparison with prior studies could be slightly more explicit (e.g., highlight how your finding that knowledge combination (β=0.42) outperforms acquisition and protection contrasts with or extends earlier KBV-based logistics studies).

Some subsections (e.g., descriptions of IoT, RFID, multimodal networks) are overly detailed and could be shortened without loss of meaning. Streamlining would keep the narrative tightly focused on theory–findings linkage.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have revised the manuscript. But some minor problems still exist.

1. The template is not stable. It could be changed. The authors may not completely use the template. So the subtitles must be added. Since each section is too long. The authors did not make efforts on this. For readers, it is very difficult to read your paper. Readers want to get the useful Information as quickly as possible by reading your subtitles and content.
2. The comparative experiment must be arranged in a separated section. Not mixed in the whole content of the method or results. The readers want to see your innovation of the work. And the comparative experiment or research is a very good way to see the innovation. The authors can not only use some comprehensive expressions to describe the comparison.
3. In each section, the secondary subtitles must be added.
4. After revising the manuscript by the above suggestions, the paper could be accepted. Otherwise, The structure of the manuscript is not suitable as to a normal paper, it can not be accepted.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 3

Reviewer 1 Report

Comments and Suggestions for Authors

The questions I raised in the previous round have already been addressed. Thus, I think the paper is now acceptable for publication in the journal.