Geotechnical Engineering: AI, Digital, Climate Resilience

Javier Ortiz^*
*Correspondence: Javier Ortiz, Department of Geotechnical Engineering, Universidad de Chile, Santiago 7500000, Chile, Email:

Introduction

The application of machine learning models in geotechnical engineering is making significant strides, particularly in predicting soil liquefaction susceptibility. By meticulously utilizing Cone Penetration Test (CPT) data, these advanced models offer a more accurate and notably efficient assessment of liquefaction risk. This improved precision is crucial for the resilient design of infrastructure, especially in seismically active areas where the consequences of liquefaction can be severe. This data-driven approach marks a pivotal advancement in mitigating geotechnical hazards [1].

Recent advancements and emerging trends in risk analysis within geotechnical engineering highlight a fundamental shift. There's a clear move towards more probabilistic and data-driven methods, which are essential for effectively managing the inherent uncertainties prevalent in ground conditions and complex engineering projects. This evolution in analytical frameworks helps engineers make more informed decisions, enhancing the reliability and safety of structures by quantitatively addressing variability and potential failures [2].

A comprehensive review illuminates the diverse and powerful applications of Artificial Intelligence (AI) techniques in geotechnical engineering. This covers a broad spectrum of areas, from enhancing site investigation processes and predicting soil properties with greater accuracy to optimizing design solutions and refining risk assessment protocols. AI's transformative potential is evident, promising to revolutionize traditional approaches and enable smarter, more efficient engineering practices across the discipline [3].

The emerging concept of Digital Twins is generating considerable discussion within geotechnical engineering, outlining both promising opportunities and significant challenges. This technology holds immense potential to enhance real-time monitoring of geotechnical assets, improve predictive modeling capabilities, and streamline the entire lifecycle management of infrastructure. By creating virtual replicas that dynamically update with physical data, Digital Twins offer unprecedented insights, fostering proactive maintenance and optimized performance throughout a project's lifespan [4].

Remote sensing and Geographic Information Systems (GIS) are proving to be invaluable for preliminary geotechnical site characterization. This review specifically focuses on their integration, explaining how these advanced technologies provide exceptionally cost-effective and efficient tools for early-stage investigations. They significantly reduce the necessity for extensive fieldwork, thereby accelerating project timelines and substantially improving the quality and speed of decision-making processes for initial site assessments [5].

A comprehensive review highlights machine learning applications specifically for predicting the engineering properties of soils. The paper covers a wide array of algorithms and discusses their demonstrated effectiveness in accurately characterizing complex soil behavior. This represents a significant move towards more predictive rather than purely empirical approaches in geotechnical practice, enhancing the scientific foundation of soil mechanics and leading to more precise and reliable engineering designs for foundations and earthworks [6].

A compelling case study demonstrates the effective use of Building Information Modeling (BIM) for geotechnical data management and visualization. BIM excels at integrating complex geotechnical information into a unified, accessible model. This capability significantly improves collaboration among various project stakeholders and enhances decision-making processes throughout a project's entire lifecycle, from planning and design to construction and maintenance. BIM, therefore, streamlines information flow and project coordination [7].

Quantifying uncertainty is a critical aspect of modern geotechnical engineering, and this review provides a thorough overview of various methods. It strongly emphasizes the paramount importance of incorporating probabilistic approaches to meticulously account for both inherent soil variability and potential measurement errors. Such rigorous methodologies are essential for developing more robust and reliable designs, ultimately contributing to safer and more resilient geotechnical structures against unpredictable ground conditions [8].

The impacts of climate change on geotechnical infrastructure are a growing concern, and this review thoughtfully explores these significant effects. It examines how altered weather patterns, rising sea levels, and an increase in extreme weather events directly affect ground conditions, structural stability, and the overall performance of infrastructure. This highlights a critical need for developing adaptive design and assessment strategies to ensure long-term resilience in the face of environmental shifts [9].

Recent advances in non-invasive geotechnical site characterization using surface wave methods are significantly improving investigative techniques. This article reviews how these methods offer efficient and considerably less disruptive alternatives compared to traditional, intrusive investigations. They provide valuable insights into subsurface conditions without the need for extensive excavation, proving essential for environmental impact mitigation and efficient project planning, making site assessments quicker and more economical [10].

Description

Geotechnical engineering is rapidly evolving, driven by advancements in data science and artificial intelligence. The application of machine learning models, particularly those leveraging Cone Penetration Test (CPT) data, offers a more accurate and efficient assessment of soil liquefaction susceptibility, which is absolutely crucial for resilient infrastructure design in seismically active areas [1]. Furthermore, a comprehensive review illustrates the diverse and powerful applications of Artificial Intelligence (AI) techniques across the field. This includes enhancing site investigation, predicting soil properties more precisely, optimizing designs, and refining risk assessment protocols, demonstrating AI's transformative potential to revolutionize traditional approaches [3]. This paradigm shift extends to comprehensive reviews highlighting machine learning's role in predicting soil engineering properties, covering various algorithms and their effectiveness in characterizing complex soil behavior. This signals a clear move towards more predictive and less empirical methods in geotechnical practice, strengthening its scientific foundation [6].

Managing uncertainty and risk is a core challenge in geotechnical engineering, and recent advancements point towards more sophisticated solutions. Emerging trends in risk analysis emphasize a shift towards probabilistic and data-driven methods, which are indispensable for effectively managing inherent uncertainties in ground conditions and complex engineering projects [2]. In support of this, various methods for quantifying uncertainty are reviewed, underscoring the critical importance of incorporating probabilistic approaches. These methods meticulously account for inherent soil variability and potential measurement errors, leading to more robust, reliable, and ultimately safer geotechnical designs against unpredictable ground conditions [8].

Innovative technologies are reshaping geotechnical site characterization and monitoring. The integration of remote sensing and Geographic Information Systems (GIS), for example, provides exceptionally cost-effective and efficient tools for preliminary site investigations. These significantly reduce the need for extensive fieldwork, thereby accelerating project timelines and improving the quality of early-stage decision-making [5]. Complementing these tools, recent advances in non-invasive geotechnical site characterization using surface wave methods offer efficient and considerably less disruptive alternatives to traditional intrusive investigations, providing valuable insights into subsurface conditions without extensive excavation [10].

Digitalization is also transforming geotechnical data management and visualization. The emerging concept of Digital Twins holds immense potential to enhance real-time monitoring of geotechnical assets, improve predictive modeling capabilities, and streamline their entire lifecycle management, despite presenting inherent challenges [4]. Similarly, a compelling case study showcases the effective use of Building Information Modeling (BIM) for geotechnical data management and visualization. BIM excels at integrating complex geotechnical information into a unified, accessible model, which significantly improves collaboration and decision-making among project stakeholders throughout a project's lifecycle [7].

Beyond project-specific applications, the geotechnical community is actively addressing broader environmental challenges. A significant review thoughtfully explores the profound impacts of climate change on geotechnical infrastructure. It meticulously examines how altered weather patterns, rising sea levels, and an increase in extreme weather events directly affect ground conditions, structural stability, and the overall performance of infrastructure [9]. This critical understanding highlights the urgent need for developing adaptive design and assessment strategies to ensure long-term resilience and safety in the face of a rapidly changing global environment.

Conclusion

Geotechnical engineering is undergoing a significant transformation, embracing advanced technologies and methodologies to address complex challenges. Machine learning and Artificial Intelligence (AI) are being widely adopted to predict soil properties, assess liquefaction susceptibility using CPT data, and optimize designs, leading to more accurate and efficient risk evaluations for infrastructure in seismically active areas. A critical shift towards probabilistic and data-driven risk analysis is evident, essential for quantifying and managing the inherent uncertainties in ground conditions and engineering projects, ensuring more robust and reliable designs. Digitalization plays a pivotal role, with technologies like Digital Twins enhancing real-time monitoring and lifecycle management of geotechnical assets, while Building Information Modeling (BIM) improves data integration, visualization, and collaboration across project phases. For site characterization, innovations include the integration of remote sensing and Geographic Information Systems (GIS) for cost-effective preliminary investigations, alongside advanced non-invasive surface wave methods that offer efficient alternatives to traditional intrusive techniques, providing vital subsurface insights. Furthermore, the field is actively confronting the impacts of climate change, examining how altered weather patterns, rising sea levels, and extreme events affect geotechnical infrastructure. This necessitates the development of adaptive design and assessment strategies to ensure long-term resilience and safety in a changing environment, pushing the discipline towards more sustainable and forward-looking practices.

Acknowledgement

None.

Conflict of Interest

None.

References

1. Mohsen R, Ramin G, Ebrahim F. "Machine learning for predicting soil liquefaction susceptibility based on CPT data".Eng Geol 295 (2021):106428.

Indexed at, Google Scholar, Crossref

2. Kok-Kwang P, Massimo U, Jianye C. "Risk analysis in geotechnical engineering: A review of recent advances and future trends".Eng Geol 269 (2020):105552.

Indexed at, Google Scholar, Crossref

3. Wei C, Chong L, Huiming H. "Application of artificial intelligence in geotechnical engineering: a comprehensive review".Geosci Front 13 (2022):101342.

Indexed at, Google Scholar, Crossref

4. Kenichi S, Nagulendran S, Maarten JD. "Digital Twin for Geotechnical Engineering: Challenges and Opportunities".Geotech Eng J SEAGS AGSSEA 52 (2021):1-13.

Indexed at, Google Scholar, Crossref

5. Rana E, Abdullah E, Yasser E. "Integration of remote sensing and GIS for preliminary geotechnical site characterization: A review".J Geophys Eng 17 (2020):746-764.

Indexed at, Google Scholar, Crossref

6. Madan K, Kushal HS, T.N. S. "Machine learning for predicting engineering properties of soils: A comprehensive review".Geosci J 24 (2020):497-512.

Indexed at, Google Scholar, Crossref

7. Seungun K, Yong-Kyu C, Hyunsoo H. "BIM-based geotechnical data management and visualization: A case study".Autom Constr 125 (2021):103607.

Indexed at, Google Scholar, Crossref

8. Jianye C, Kok-Kwang P, Massimo U. "Uncertainty quantification in geotechnical engineering: A review".Geotech Eng J SEAGS AGSSEA 50 (2019):1-22.

Indexed at, Google Scholar, Crossref

9. Omid R, Haryati M, Alireza S. "Impacts of climate change on geotechnical infrastructure: A review".Environ Sci Pollut Res 30 (2023):73602-73623.

Indexed at, Google Scholar, Crossref

10. Sebastiano F, Jean-Sébastien L, Francesco G. "Recent advances in non-invasive geotechnical site characterization using surface waves".Geotech Eng J SEAGS AGSSEA 51 (2020):1-21.

Indexed at, Google Scholar, Crossref