Artificial Intelligence Applications in Soil Compaction Engineering

Artificial intelligence (AI) has emerged as a transformative tool in soil compaction engineering, enabling the rapid prediction of key parameters that traditionally require time-consuming laboratory tests. By leveraging data-driven models, engineers can estimate optimum moisture content, maximum dry density and bearing ratios from readily obtained soil properties, reducing field uncertainty and speeding up decision-making. Machine learning techniques such as artificial neural networks, ensemble methods and hybrid intelligence frameworks have been applied to diverse soil types—from fine-grained clays to amended mixtures—demonstrating strong predictive performance across global datasets. These approaches support sustainable construction by minimising material waste and carbon emissions associated with repeated testing, while enhancing quality control through real-time compaction monitoring. As digital sensing and data acquisition in earthworks become more widespread, AI-based models promise to integrate with automated rollers and control systems, ushering in a new era of precision geotechnical practice.

Research from Nature Portfolio

No recent Nature Portfolio content available.

Research from all publishers

In 2024, a comprehensive review consolidated correlations between geotechnical soil properties and California Bearing Ratio (CBR), highlighting how machine learning regression models can predict CBR from index and compaction parameters, thereby guiding pavement design more efficiently. A 2022 study applied optimisable ensemble algorithms alongside artificial neural networks to forecast optimum moisture content, maximum dry density and unconfined compressive strength of soil amendments drawn from a global dataset, achieving high accuracy and demonstrating the transferability of AI models across regions. More recently, a hybrid intelligence paradigm combining an adaptive neuro-fuzzy inference system with an improved grey wolf optimiser achieved precise estimations of compaction parameters, offering a robust computational alternative to conventional laboratory tests and enabling potential real-time control of compaction equipment on site.

Artificial Intelligence Applications in Soil Compaction Engineering publication trend

The graph below shows the total number of articles in artificial intelligence applications in soil compaction engineering across all publications each year (not limited to Nature Index journals).

Technical terms

Optimum moisture content (OMC): The water content at which soil attains its maximum dry density under a specified compactive effort.

Maximum dry density (MDD): The highest density achieved by a soil sample at its OMC during laboratory compaction tests.

California Bearing Ratio (CBR): A penetration-based index evaluating the strength and stiffness of subgrade soils and pavement materials.

Artificial Neural Network (ANN): A computational model inspired by the human brain’s network of neurons, used to learn complex, nonlinear relationships from data.

Adaptive Neuro-Fuzzy Inference System (ANFIS): A hybrid modelling approach that combines neural network learning capabilities with fuzzy logic rule-based systems.

Grey Wolf Optimiser (GWO): A nature-inspired metaheuristic algorithm that mimics grey wolf social hierarchy and hunting behaviour for solution optimisation.

References

1. The effect of geotechnical soil properties on cbr value: review. AI in Civil Engineering (2024).
2. Prediction of Compaction and Strength Properties of Amended Soil Using Machine Learning. Buildings (2022).
3. Modelling Soil Compaction Parameters Using an Enhanced Hybrid Intelligence Paradigm of ANFIS and Improved Grey Wolf Optimiser. Mathematics (2023).