2510.0012 A Review of Intelligent Rock Mechanics: From Methods to Applications v1

🎯 ICAIS2025 Submission
Back to Paper List

Paper Content

📄 Open in New Tab

AI Review from DeepReviewer

AI Review available after:
--d --h --m --s

📋 AI Review from DeepReviewer will be automatically processed

📋 Summary

This paper provides a comprehensive review of the integration of artificial intelligence (AI) and machine learning (ML) techniques into the field of rock mechanics, a domain traditionally reliant on empirical methods and numerical modeling. The authors effectively trace the evolution of AI applications in this field, starting from early methods like backpropagation and support vector machines to modern deep learning architectures such as convolutional neural networks (CNNs), long short-term memory networks (LSTMs), and generative adversarial networks (GANs). The core contribution of this work lies in its synthesis of the current state of 'intelligent rock mechanics,' highlighting how AI can address long-standing challenges in the field, including the complexities introduced by anisotropy, discontinuities, and multi-physics interactions inherent in geological materials. The paper is structured around several key themes, including data-driven estimation of rock properties, image-based modeling and fracture detection, AI-assisted constitutive modeling and simulation, and applications in rock engineering. In the data-driven estimation section, the authors discuss how AI is used to predict rock properties like uniaxial compressive strength (UCS) and elastic modulus from various types of input data, such as scanline data and non-destructive tests. They highlight the use of machine learning models to learn complex relationships between input features and rock properties, often achieving higher accuracy than traditional empirical methods. The section on image-based modeling focuses on the use of AI, particularly CNNs, for automating the detection and characterization of fractures in rock masses from images and scanline data. This is a crucial area, as accurate fracture characterization is essential for understanding rock mass behavior. The paper also explores the emerging field of AI-assisted constitutive modeling, where AI is used to enhance traditional constitutive models or develop new ones. This includes the use of physics-informed neural networks (PINNs) to incorporate physical constraints into the learning process, ensuring that the models are not only data-driven but also physically meaningful. Finally, the paper discusses various applications of AI in rock engineering, such as rockburst prediction, slope stability analysis, and tunneling. These applications demonstrate the potential of AI to improve the efficiency and accuracy of engineering decisions in rock mechanics. Overall, the paper serves as a valuable resource for researchers and practitioners interested in the intersection of AI and rock mechanics. It provides a broad overview of the current state of the field, highlighting key advancements and identifying areas where further research is needed. The authors emphasize the potential of AI to revolutionize rock mechanics by providing new tools for data-driven modeling, real-time hazard prediction, and the integration of diverse data sources. However, they also acknowledge the challenges that need to be overcome, such as data scarcity, model generalization, and the need for interdisciplinary collaboration. The paper concludes by outlining future research directions, including the development of standardized datasets, the integration of domain knowledge into AI models, and the exploration of new AI techniques for addressing complex rock mechanics problems. While the paper does not present novel experimental results or theoretical insights, its value lies in its comprehensive synthesis of the existing literature and its ability to provide a clear and accessible overview of a rapidly evolving field. It effectively communicates the potential of AI to enhance traditional rock mechanics methods and offers a roadmap for future research and development in this exciting interdisciplinary area.

✅ Strengths

One of the paper's primary strengths is its comprehensive literature review. I appreciate the authors' effort to synthesize a wide range of AI techniques and their applications in rock mechanics. The paper effectively covers the evolution of AI in this field, from early methods to modern deep learning approaches. I find the organization of the review into thematic sections particularly helpful. The sections on data-driven property estimation, image-based modeling, AI-assisted constitutive modeling, and applications in rock engineering provide a clear and logical structure that makes it easy to follow the development of the field. The authors do an excellent job of highlighting the key advancements in each area and providing specific examples of how AI is being used to address practical problems in rock mechanics. For instance, the discussion of CNNs for fracture detection and PINNs for constitutive modeling demonstrates a deep understanding of the current state-of-the-art. I also appreciate the paper's emphasis on the interdisciplinary nature of this research area. The authors effectively highlight the convergence of AI and rock mechanics, showcasing how AI can enhance traditional methods and offer new tools for data-driven modeling and real-time hazard prediction. This interdisciplinary perspective is crucial for advancing the field and fostering collaboration between AI experts and rock mechanics specialists. Furthermore, the paper's discussion of emerging techniques, such as physics-informed neural networks and graph-based learning, demonstrates a forward-looking perspective. I believe the authors accurately identify the potential of these techniques to bridge the gap between data-driven models and physical interpretability, which is a critical challenge in applying AI to scientific domains. The paper also provides a valuable service by outlining future research directions. The emphasis on the need for standardized datasets, interdisciplinary collaboration, and the integration of domain knowledge into AI models is particularly important. I agree with the authors that addressing these challenges is crucial for realizing the full potential of AI in rock mechanics. Finally, the paper is generally well-written and accessible, even for readers who may not be experts in both AI and rock mechanics. The authors provide clear explanations of complex concepts and avoid excessive jargon, making the paper engaging and informative.

❌ Weaknesses

Despite its strengths, I find that the paper has several limitations that warrant careful consideration. One of the most significant weaknesses is the lack of concrete, real-world examples that demonstrate the practical application of AI techniques in rock mechanics projects. While the paper provides numerous examples of studies and their findings, these often lack the depth and detail needed to fully understand how AI is being applied in practice. For instance, in the section on data-driven estimation of rock properties, the paper mentions an ANN model for predicting UCS but doesn't provide specifics about the project context, the data acquisition process, or the engineering implications of the model's predictions. This lack of detail makes it difficult to assess the true impact of these methods on real-world engineering practice. The paper would be significantly more impactful if it included in-depth case studies that showcased the application of AI in specific rock mechanics projects. These case studies should detail the problem statement, the data used, the AI methods employed, the challenges encountered, and the results obtained. Without such concrete examples, the paper risks being perceived as a theoretical overview rather than a practical guide for practitioners. Another major limitation is the limited discussion on model generalization. The paper acknowledges the challenge of generalization, stating that "Many AI models discussed are trained on relatively small datasets or data from specific sites, which raises concerns about overfitting and generalizability." However, it fails to delve deeply into strategies for improving the robustness and transferability of AI models across different geological settings. This is a critical issue in applying AI to rock mechanics, where data scarcity and heterogeneity are common. The paper should have explored techniques such as domain adaptation, transfer learning, and ensemble methods in more detail. For example, it could have discussed how pre-trained models on large, diverse datasets could be fine-tuned for specific applications or how models trained on data from one geological setting could be adapted to another. The absence of a thorough discussion on generalization strategies significantly undermines the paper's practical relevance, as it leaves the reader with unanswered questions about how to ensure the reliability of AI models in diverse real-world scenarios. Furthermore, the paper does not adequately address the challenge of data quality and availability. While it acknowledges data scarcity as a significant challenge, it does not explore potential solutions for addressing this issue in detail. The paper should have discussed methods for data augmentation, synthetic data generation, and collaborative data-sharing initiatives. For instance, it could have explored the use of generative adversarial networks (GANs) for generating realistic synthetic rock images or the use of physics-based simulations to create synthetic datasets for training AI models. Additionally, the paper should have emphasized the importance of collaborative data-sharing initiatives, where researchers and practitioners can share their data to create larger, more diverse datasets for training AI models. The lack of a detailed discussion on these solutions limits the paper's ability to provide practical guidance on overcoming one of the most significant hurdles in applying AI to rock mechanics. The paper also lacks a critical analysis of the limitations of current AI methods in rock mechanics. While it briefly mentions some challenges in the "CHALLENGES AND FRONTIER DISCUSSIONS" section, it does not delve into the inherent limitations of these methods. For example, the paper does not adequately discuss the lack of interpretability of many deep learning models, which can make it difficult to understand the underlying physical mechanisms driving their predictions. This is a crucial issue in a field like rock mechanics, where understanding the "why" behind a prediction is often as important as the prediction itself. Additionally, the paper does not thoroughly address the sensitivity of AI models to hyperparameter tuning and the potential for overfitting, especially when dealing with limited datasets. A more critical analysis of these limitations would have provided a more balanced perspective on the current state of AI in rock mechanics and highlighted areas where further research is needed. Finally, while the paper provides a comprehensive overview of the current state of AI in rock mechanics, it does not offer any new insights or findings. Its primary contribution is synthesizing existing knowledge, which, while valuable, limits its novelty. The paper does not present any novel experimental results, theoretical advancements, or innovative methodologies. This lack of novelty is a significant limitation, especially considering the rapid pace of advancements in both AI and rock mechanics. To be more impactful, the paper could have gone beyond summarizing existing literature and proposed new research directions or innovative applications of AI in this field. In summary, the paper's weaknesses lie in its lack of concrete examples, limited discussion on model generalization and data scarcity solutions, insufficient critical analysis of AI limitations, and lack of novel contributions. These limitations, which I have identified and validated through a thorough examination of the paper's content, significantly impact the paper's overall practical relevance and usefulness to both researchers and practitioners in the field. My confidence in these identified weaknesses is high, as they are consistently supported by the paper's content and structure.

💡 Suggestions

To enhance the practical relevance and impact of this review, I would strongly recommend the inclusion of detailed case studies that illustrate the application of AI techniques in real-world rock mechanics projects. These case studies should go beyond simply summarizing existing research and provide in-depth narratives of how AI is being applied to solve specific engineering problems. For example, a case study could focus on the use of convolutional neural networks (CNNs) for automated rock mass classification from scanline data in a specific tunneling project. This case study should detail the data acquisition process, including the types of sensors used and the data preprocessing steps. It should also describe the specific CNN architecture employed, the training process, and the performance metrics used to evaluate the model's accuracy. Furthermore, it should discuss the challenges encountered during the implementation, such as data quality issues or computational limitations, and how these challenges were addressed. Finally, it should analyze the impact of the AI model on the project, such as improvements in efficiency, accuracy, or safety. Another valuable case study could explore the use of physics-informed neural networks (PINNs) for simulating rock deformation under different stress conditions in an underground mining operation. This case study should detail how the PINN model was integrated with existing numerical methods, such as finite element analysis, and how it improved the accuracy of predictions compared to traditional approaches. It should also discuss the specific benefits of using PINNs, such as their ability to incorporate physical constraints and their potential for real-time monitoring and prediction. These concrete examples would not only make the review more engaging but also provide valuable guidance for practitioners seeking to implement AI in their work. In addition to case studies, the review should delve deeper into the issue of model generalization. I recommend including a dedicated section that explores specific techniques for improving the robustness and transferability of AI models across different geological settings. This section should discuss domain adaptation methods, which aim to adapt models trained on one domain to perform well on another. For example, it could explore techniques such as transfer learning, where a model pre-trained on a large, diverse dataset is fine-tuned on a smaller, domain-specific dataset. The review should also discuss the importance of data normalization and standardization in improving model generalization, as well as the use of ensemble methods to combine the predictions of multiple models trained on different datasets. A detailed analysis of these techniques, along with examples of their application in rock mechanics, would significantly enhance the practical value of the review. Furthermore, the review should address the challenge of data scarcity by exploring potential solutions for data augmentation and synthetic data generation. I suggest including a section that discusses techniques such as generative adversarial networks (GANs) for generating realistic synthetic rock images or the use of physics-based simulations to create synthetic datasets for training AI models. The review should also emphasize the importance of collaborative data-sharing initiatives, where researchers and practitioners can share their data to create larger, more diverse datasets for training AI models. This would not only help to address the data scarcity problem but also promote the development of more robust and generalizable AI models for rock mechanics. To provide a more balanced perspective, the review should include a critical analysis of the limitations of current AI methods in rock mechanics. I recommend discussing the lack of interpretability of many deep learning models and the challenges this poses for understanding the underlying physical mechanisms driving their predictions. The review should also address the sensitivity of AI models to hyperparameter tuning and the potential for overfitting, especially when dealing with limited datasets. This critical analysis would help to set realistic expectations for the capabilities of AI in rock mechanics and highlight areas where further research is needed. Finally, to enhance the novelty and impact of the review, I suggest including a section that proposes new research directions or innovative applications of AI in rock mechanics. This could involve identifying areas where AI is likely to have the greatest impact, such as integrating different types of data (e.g., geological, geophysical, and geotechnical data) to create more comprehensive models of rock mass behavior. It could also involve discussing the potential of using AI to develop more efficient and robust methods for data acquisition and processing. By proposing new research directions, the review can contribute to the advancement of the field and inspire further innovation in the application of AI to rock mechanics.

❓ Questions

Given the paper's focus on the application of AI in rock mechanics, I am curious about the specific challenges encountered when integrating AI models with existing engineering workflows. How do practitioners typically incorporate AI predictions into their decision-making processes, and what are the main hurdles in this integration? I am also interested in understanding the level of acceptance of AI-based methods within the rock mechanics community. Are there any studies or surveys that shed light on the willingness of engineers to adopt AI in their practice, and what are the main concerns or reservations they may have? Regarding data quality and availability, I wonder what the authors' perspective is on the trade-offs between using limited, site-specific data versus more general, but potentially less accurate, models. In scenarios where data is scarce, is it better to rely on simpler, more interpretable models trained on-site, or to use more complex models trained on larger, external datasets with potential domain shifts? I am also curious about the authors' thoughts on the interpretability of AI models in rock mechanics. Given the importance of understanding the underlying physical mechanisms in this field, how do they see the trade-off between model accuracy and interpretability evolving in the future? Do they foresee a shift towards more interpretable models, or will the focus remain on achieving the highest possible accuracy, even if the models are less transparent? Furthermore, I would like to know more about the authors' views on the role of domain knowledge in AI model development. How can we effectively integrate geological and geotechnical expertise into AI models to improve their performance and reliability? Are there any specific approaches or frameworks that the authors believe are particularly promising in this regard? Finally, considering the rapid advancements in AI, what are the authors' thoughts on the future of AI in rock mechanics over the next 5-10 years? Which areas do they believe are most likely to see significant breakthroughs, and what are the key research questions that need to be addressed to fully realize the potential of AI in this field? I am particularly interested in their perspective on the potential of emerging AI techniques, such as graph neural networks or reinforcement learning, to address complex rock mechanics problems.

📊 Scores

Soundness:2.75
Presentation:3.0
Contribution:2.5
Rating: 5.75

AI Review from ZGCA

ZGCA Review available after:
--d --h --m --s

📋 AI Review from ZGCA will be automatically processed

📋 Summary

This is a narrative review of intelligent rock mechanics, synthesizing AI/ML methods and applications in rock engineering. The paper surveys: (i) data-driven estimation of rock properties (e.g., UCS, elastic moduli, wave velocities) using ANNs, SVMs, ANFIS, ensembles, and Bayesian frameworks (Section 2); (ii) image-based modeling and fracture detection, including deep learning for 3D digital rock reconstruction (VAEs, GANs), and CNN-based fracture segmentation from images, CT, LiDAR/photogrammetry point clouds, and transfer learning for DFN mapping (Section 3); (iii) AI-assisted constitutive modeling and simulation with PINNs and variants (enriched PINNs, PIRBNs, PI-TCN), multiscale surrogates, GNNs for elasticity surrogates, RNNs for phase-field fracture emulation, and physics-embedded operators (Section 4); and (iv) applications: rock mass classification (RMR/Q/GSI, drill-core RQD), rockburst and geohazard prediction (ensembles on historical and monitoring data), tunneling/TBM operations (deformation prediction, RL control, vision-based muck analysis), and slope stability (classification, reliability with ML surrogates) (Section 5). Section 6 discusses challenges—data scarcity/standardization, interpretability/trust (including hybrid physics-based learning), computational efficiency, workflow integration, and model maintenance—and projects future directions (e.g., PINNs in routine design, digital twins, and exploratory roles for LLMs).

✅ Strengths

  • Broad coverage of AI techniques relevant to rock mechanics, spanning property prediction, image-based micro/macro characterization, physics-informed modeling, and several real-world application domains (Sections 2–5).
  • Provides nuanced discussion of trade-offs and limitations across methods, e.g., reconstruction via simulated annealing vs deep 3D models, PINN stability and training curricula, ensemble benefits for rockburst prediction, and domain shift issues in fracture segmentation (Sections 3–4).
  • Highlights hybrid physics–ML approaches (PINNs, PIRBNs, physics-embedded operators, multiscale surrogates) and their promise for bridging data-driven models with mechanistic constraints (Section 4; Section 6).
  • Identifies key cross-cutting challenges—data quality/standardization, interpretability and uncertainty, computational cost, workflow integration, and lifecycle maintenance—and points to concrete mitigation strategies (Section 6).
  • Forward-looking perspective tying newer ML paradigms (e.g., LLM-enabled tooling and decision support) to potential geotechnical workflows (Section 6).

❌ Weaknesses

  • Lack of a systematic review methodology: no search protocol, inclusion/exclusion criteria, time window, or coverage bounds; no quality assessment of included studies. This limits reproducibility and makes coverage completeness unclear.
  • No visual synthesis (taxonomy, summary tables of tasks/datasets/metrics, comparative matrices, or method schematics). As a result, readers cannot easily compare methods, datasets, benchmarks, or performance across application areas, which undermines the stated goal of a structured, forward-looking roadmap.
  • Citation quality and alignment issues: several references appear mismatched or inappropriately attributed (e.g., McCulloch & Pitts, 1943 in a rock mechanics context in Section 1; future-dated references such as Li et al., 2025; apparent typos and cross-domain misplacements in the bibliography). These issues weaken the scholarly foundation.
  • Limited critical comparison of state-of-the-art performance: few concrete, side-by-side evaluations of methods on shared datasets or benchmarks; limited discussion of failure modes (e.g., OOD generalization in fracture mapping, robustness of hazard predictors), and scarce quantitative consolidation.
  • Claims regarding LLMs and automated code generation/decision support in geotechnical practice are largely speculative here and not grounded with concrete, domain-specific case studies or references (Section 6).
  • No explicit ethics/impact section tailored to safety-critical geotechnical decision-making (e.g., risk of false positives/negatives in hazard prediction, automation bias, accountability), despite discussing trust and interpretability.

❓ Questions

  • Can you describe the review protocol in detail (databases searched, query strings, time window, inclusion/exclusion criteria, study selection flow, quality assessment), and provide a PRISMA-style summary to establish coverage and reproducibility?
  • Please add a taxonomy figure that organizes methods (data-driven, image-based, physics-informed, hybrid) against rock mechanics tasks (properties, microstructure, constitutive modeling, applications) and data regimes (lab/field, image/time-series/point cloud), with pointers to canonical references.
  • Could you include comparative tables summarizing, for each application area, key datasets (size, modality, labeling), methods, evaluation metrics, and best-reported performance, highlighting limitations (e.g., site specificity, OOD performance, uncertainty reporting)?
  • Several references seem misaligned or future-dated. Can you audit and correct citations so each claim is supported by appropriate, domain-relevant sources? For instance, clarify the role of classic AI citations in a rock mechanics context (Section 1) and ensure all application claims map to concrete rock/geotechnical studies.
  • For PINNs and physics-informed variants, can you provide a concise comparison of training strategies (e.g., non-dimensionalization, curriculum/sequential training, residual weighting) and typical failure modes on geomechanics benchmarks (e.g., Mandel’s problem, poroelasticity) with references?
  • For fracture detection and DFN reconstruction, how do methods generalize across rock types, imaging conditions, and sensors? Can you discuss domain adaptation and uncertainty quantification strategies, with examples?
  • Your discussion of LLMs is forward-looking. Can you ground it with concrete geotechnical use cases (e.g., code synthesis for FEM pre/post-processing, report summarization with retrieval augmentation, or decision support with safety guardrails), and discuss risks (hallucination, traceability)?
  • Could you include an ethics/impact subsection covering safety-critical deployment (alarm thresholds, human-in-the-loop, documentation of uncertainty), dataset bias (site/rock-type imbalance), and governance for model updates over project lifecycles?
  • Will you provide a curated list of open datasets, benchmark problems, and baseline implementations to facilitate reproducible progress in intelligent rock mechanics?

⚠️ Limitations

  • The review acknowledges data scarcity, generalization, and interpretability challenges (Section 6), but does not provide a structured resource (datasets/benchmarks) to mitigate them.
  • Lack of systematic methodology and absence of comparative tables hinder reproducibility of the review and clarity of the synthesis.
  • Potential negative societal impacts—stemming from safety-critical mispredictions in hazard assessment, automation bias in decision support, and opaque models—are not explicitly treated. High-stakes contexts (rockburst early warning, slope instability) require uncertainty quantification, conservative deployment strategies, and human oversight.
  • Speculative discussion on LLMs lacks cautionary guidance about hallucinations, provenance tracking, and validation in engineering workflows.
  • Maintenance and drift management are noted (Section 6) but would benefit from concrete protocols (data/version governance, validation gates, recalibration schedules).

🖼️ Image Evaluation

Cross-Modal Consistency: 44/50

Textual Logical Soundness: 12/30

Visual Aesthetics & Clarity: 14/20

Overall Score: 70/100

Detailed Evaluation (≤500 words):

1. Cross-Modal Consistency

• Major 1: Missing internal equation reference; the text points to a non-existent Eq. 2. Evidence: Sec 5.5: “fragment size distribution … (\ref{eq:2})”.

• Minor 1: No figures/tables provided; difficult to align methods/performance claims with visuals. Evidence: Entire manuscript lacks figures/tables.

• Minor 2: Occasional ambiguous attributions to specific models without clear identifiers (e.g., “FraSegNet”) complicate verification. Evidence: Sec 3: “FraSegNet … (Lee et al., 2022)”.

2. Text Logic

• Major 1: Miscitation of foundational works for specific empirical claims (UCS prediction from 108 samples). Evidence: Sec 2: “108 samples … (Abueidda et al., 2021; Turing, 1950)”.

• Major 2: Incorrect reference for CNN microseismic localization attributed to LSTM paper. Evidence: Sec 1: “CNNs can localize microseismic … (Hochreiter & Schmidhuber, 1997)”.

• Major 3: RQD accuracy claim mis-cited; the referenced paper is about rockburst risk, not RQD. Evidence: Sec 5.1: “CNNs … predicted RQD within 2–3% … (Guo et al., 2022)”.

• Minor 1: FraSegNet citation unclear/mismatched. Evidence: Sec 3: “FraSegNet … (Lee et al., 2022)” while refs show Lee 2022 on elastic constants.

• Minor 2: Anachronistic/irrelevant citations in key places. Evidence: Sec 1: “(McCulloch & Pitts, 1943)” used to motivate rock mechanics challenges.

3. Figure Quality

• Major 1: Absence of schematics/summary tables hampers quick grasp of taxonomy, datasets, and comparisons; Figure‑Alone test not possible. Evidence: No figures/tables present in manuscript.

• Minor 1: Visible LaTeX placeholder harms readability. Evidence: Sec 5.5: “(\ref{eq:2})”.

Key strengths:

  • Broad, timely survey spanning properties, imaging, PINNs/GNNs, and applications (tunneling, rockburst, slopes).
  • Clear high-level structure from methods to applications and challenges.
  • Useful discussion on hybrid physics-informed approaches and workflow integration.

Key weaknesses:

  • Pervasive citation/attribution errors (multiple concrete examples above) undermine credibility of claims.
  • No visuals or summary tables; add at least: (a) taxonomy diagram of methods vs. tasks; (b) table mapping tasks→datasets→models→metrics; (c) workflow schematic for hybrid PINN/FEM pipelines.
  • Internal cross-reference error to Eq. 2; resolve or remove.
  • Some over-specific performance claims without verifiable sources; quantify with consistent metrics and correct citations, or soften.

📊 Scores

Originality:3
Quality:2
Clarity:2
Significance:2
Soundness:2
Presentation:2
Contribution:2
Rating: 4

AI Review from SafeReviewer

SafeReviewer Review available after:
--d --h --m --s

📋 AI Review from SafeReviewer will be automatically processed

📋 Summary

The manuscript provides a review on the applications of artificial intelligence (AI) and machine learning (ML) in the field of rock mechanics. It covers a wide range of topics, from fundamental methods to practical applications. The paper highlights the evolution of AI in this field, from early techniques to modern approaches such as deep learning frameworks and physics-informed neural networks. The authors also discuss the challenges and future directions for this emerging interdisciplinary field.

✅ Strengths

The paper provides a broad overview of AI applications in rock mechanics, covering various techniques and applications. The writing is generally clear and easy to follow, and the paper includes a discussion of challenges and future directions.

❌ Weaknesses

The paper's broad scope limits the depth of discussion for each topic. It lacks detailed explanations of the AI methods and the fundamental principles of rock mechanics. The paper also lacks specific examples and quantitative results to support its claims. The discussion of challenges and future directions is somewhat general and could be more specific and actionable. The paper does not explicitly define its target audience.

💡 Suggestions

To improve the paper, I would suggest focusing on a more specific aspect of AI in rock mechanics or providing more in-depth analysis of the current topics. Including detailed examples of successful AI applications, along with quantitative results, would significantly enhance the paper's value. The authors should also consider adding a section that explicitly defines the target audience and tailors the content accordingly. Furthermore, the discussion of challenges and future directions could be more specific, providing concrete recommendations for future research.

❓ Questions

What is the intended audience for this review paper? Are there specific areas where the authors felt they had to make significant compromises in terms of depth due to the broad scope? Can the authors provide examples of specific research gaps that they believe are most critical for advancing the field of AI in rock mechanics?

📊 Scores

Soundness:3
Presentation:3
Contribution:3
Confidence:3.0
Rating: 5.5

Keywords

Click the button to extract keywords

Insights

Click the button to extract insights
Version Management
Version 1 ⚠️ Not latest
📋 Version History
Citation Tools

📝 Cite This Paper