Combining slope satellite image analysis and artificial intelligence algorithms for highway service level assessment

1. Home
2. Publications
3. Online Library
4. Combining slope satellite image analysis and artificial intelligence algorithms for highway service level assessment

Publication Title
Combining slope satellite image analysis and artificial intelligence algorithms for highway service level assessment

Abstract
The study proposes an integrated approach using Artificial Intelligence (AI), satellite imagery analysis, and Machine Learning (ML) to address the global challenge of slope instabilities. It emphasizes the importance of comprehensive risk assessment frameworks, particularly focusing on Early Warning Systems (EWS) and decision-making tools. By structuring risk assessment around the hazard-exposure-vulnerability-damage model, the study aims to provide a systematic approach to slope instability management, especially in transportation networks. Additionally, the research evaluates ML algorithms, namely, Multiple linear Regression (MR) and Artificial Neural Networks (ANN), for displacement prediction, highlighting their promising performance while identifying areas for refinement in data preprocessing and model optimization. MR and ANN models consistently achieved high performance, with determinant coefficient (R2) values of 0.93. This indicates that the model can explain approximately 93% of the variance in the target variable, reflecting strong predictive capability. Despite minor variations in other metrics, such as Mean Absolute Error, Mean Squared Error, and Root Mean Squared Error, the R2 values remained consistent, emphasizing the robustness of the model in predicting displacements.
Authors
D. Owusu-Ansah; J. Tinoco; J. Matos; R. Cabral

Published In

18th European Conference on Soil Mechanics and Geotechnical Engineering (ECSMGE2024)

Theme

B - Geohazards

Download PDF