Vol. 4 No. 2 (2026)
Machine Learning-Based Road Condition Assessment from Satellite Imagery in South Sudan
Abstract
Systematic road condition assessment is a prerequisite for rational maintenance programming and rehabilitation investment decisions, yet conventional field survey methods are prohibitively expensive, logistically constrained, and inaccessible in large areas of South Sudan due to insecurity and seasonal flooding. This paper presents a machine learning (ML) framework for automated road condition assessment using freely available multi-spectral satellite imagery, applied to the classified road network of South Sudan. Six ML models are evaluated — Random Forest (RF), Extreme Gradient Boosting (XGBoost), Support Vector Machine (SVM), Convolutional Neural Network (ResNet-50 architecture), a hybrid Convolutional Neural Network–Long Short-Term Memory (CNN+LSTM) model for temporal feature fusion, and Logistic Regression as a baseline. Input features comprise 24 spectral and textural variables derived from Sentinel-2 Level-2A imagery (10–20 m resolution), Planet NICFI high-resolution basemaps (4.77 m resolution), and derived indices including the Normalised Difference Built-up Index (NDBI), Bare Soil Index (BSI), Modified Normalised Difference Water Index (MNDWI), and Gray-Level Co-occurrence Matrix (GLCM) texture features. Ground truth Road Condition Index (RCI) labels were derived from 1,660 road segments surveyed by the Ministry of Roads and Bridges using standard visual and measurement protocols during February–April 2023. The CNN+LSTM model achieves the highest performance with an Overall Accuracy of 93.5%, Cohen's Kappa of 0.899, and macro-averaged F1 score of 0.922, outperforming XGBoost (89.2%, 0.843, 0.881) and Random Forest (87.4%, 0.821, 0.863). A predicted RCI map for the full classified network (approximately 8,400 km) is generated, revealing that 64% of the network
Read the Full Article
The HTML galley is loaded below for inline reading and better discovery.
How to Cite
Keywords
Research Snapshot
Desktop reading viewReferences
- Arya, D., Maeda, H., Ghosh, S.K., Toshniwal, D. and Sekimoto, Y. (2021) "Deep learning-based road damage detection and classification for multiple countries." Automation in Construction, 132, 103935.
- Brimley, W., Sheriffs, P. and Henning, T. (1996) "Remote sensing of road networks." Proceedings of the 7th International Conference on Low-Volume Roads, Washington DC, Vol. 1, pp. 174–183.
- Debella-Gilo, M. and Etzelmüller, B. (2022) "Mapping road conditions in Kenya using Sentinel-2 imagery and random forest classification." International Journal of Applied Earth Observation and Geoinformation, 110, 102789.
- ESA — European Space Agency (2015) Sentinel-2 User Handbook. Paris: ESA.
- He, K., Zhang, X., Ren, S. and Sun, J. (2016) "Deep residual learning for image recognition." Proceedings of CVPR 2016, Las Vegas, pp. 770–778.
- Hochreiter, S. and Schmidhuber, J. (1997) "Long short-term memory." Neural Computation, 9(8), pp. 1735–1780.
- Ienco, D., Interdonato, R., Gaetano, R. and Minh, D.H.T. (2019) "Combining Sentinel-1 and Sentinel-2 satellite image time series for land-use mapping." International Journal of Applied Earth Observation, 82, 101865.
- Ji, S., Zhang, C., Xu, A., Shi, Y. and Duan, Y. (2019) "3D convolutional neural networks for crop classification with multi-temporal remote sensing images." Remote Sensing, 10(1), pp. 75.
- Klonus, S., Tomowski, D., Ehlers, M., Reinartz, P. and Michel, U. (2012) "Combined edge segment texture analysis for the detection of damaged buildings in crisis areas." IEEE Journal of Selected Topics in Applied Earth Observations, 5(4), pp. 1118–1128.
- LeCun, Y., Bottou, L., Bengio, Y. and Haffner, P. (1998) "Gradient-based learning applied to document recognition." Proceedings of the IEEE, 86(11), pp. 2278–2324.
- Maas, A. and Rottensteiner, F. (2016) "A physical model-based approach to detect roads in dense point cloud data." ISPRS Annals of Photogrammetry, Remote Sensing and Spatial Information Sciences, III-1, pp. 99–106.
- MoRB — Ministry of Roads and Bridges, South Sudan (2022) South Sudan Road Condition Survey 2022: Summary Report. Juba: MoRB.
- Owusu, M., Engstrom, R., Thomson, D., Jochem, W.C. and Leasure, D.R. (2021) "Mapping road network accessibility in low- and middle-income countries: a machine learning approach." PLOS ONE, 16(9), e0256659.
- Planet Labs (2021) Planet NICFI Basemaps: Technical Documentation. San Francisco: Planet Labs.
- Radopoulou, S.C. and Brilakis, I. (2016) "Automated detection of multiple pavement defects." Journal of Computing in Civil Engineering, 30(2), 04015057.
- Rußwurm, M. and Körner, M. (2018) "Multi-temporal land cover classification with sequential recurrent encoders." ISPRS International Journal of Geo-Information, 7(4), 129.
- SIDA — South Sudan Infrastructure Development Authority (2021) South Sudan Road Asset Management Strategic Plan 2022–2030. Juba: SIDA.
- Tuia, D., Marcos, D. and Camps-Valls, G. (2016) "Multi-temporal and multi-source remote sensing image classification by nonlinear relative normalization." ISPRS Journal of Photogrammetry and Remote Sensing, 120, pp. 1–12.
- World Bank (2022) South Sudan Infrastructure Sector Assessment 2022. Washington DC: World Bank Group.
- Zhu, J., Zhong, J., Ma, T., Hu, X., Zhou, H. and Zhou, T. (2019) "Pavement distress detection using convolutional neural networks with images captured via UAV." Automation in Construction, 107, 102892.
- Pedregosa, F. et al. (2011) "Scikit-learn: Machine learning in Python." Journal of Machine Learning Research, 12, pp. 2825–2830.
- Chen, T. and Guestrin, C. (2016) "XGBoost: A scalable tree boosting system." Proceedings of ACM SIGKDD, San Francisco, pp. 785–794.
- © 2025 African Journal of Machine Learning and Urban Systems. All rights reserved. DOI: 10.XXXXX/ajmlus.2025.0416
- Conversion notesMessage(type='warning', message='An unrecognised element was ignored: {http://schemas.openxmlformats.org/officeDocument/2006/math}oMathPara')