Urban traffic congestion in Port Harcourt, Rivers State, Nigeria, remains a critical infrastructural challenge, with average peak-hour vehicle speeds on key corridors falling below 5 km/h at several major intersections. This study presents a Graph Neural Network-based Route Optimization and Recommendation System (GNN-RORS) applied to four interconnected urban road corridors such as Ikwerre Road, Government Reserved Area (GRA), Ada George Road, and Iwofe Road. The road network is formalized as a directed weighted graph where intersections are nodes and road segments are dynamic edges. A hybrid architecture combining Graph Attention Networks (GAT) for spatial encoding with Long Short-Term Memory (LSTM) networks for temporal encoding is proposed to simultaneously capture cross-corridor traffic dependencies and time-varying congestion patterns. Furthermore, traffic data collected across 28 days at seven monitored intersections yield a clean dataset of 7,627 records for model training, validation, and testing. The proposed model achieves a Mean Absolute Error (MAE) of 2.34 vehicles/min, outperforming five established baselines including DCRNN (MAE 3.56), STGCN (MAE 3.12), and standalone LSTM (MAE 4.81). Route recommendations generated from predicted edge-weight graphs reduce average journey times by 31.2% across five major origin-destination pairs. The congestion heatmap analysis confirms that GNN-based re-routing redistributes traffic load more evenly across corridors, reducing peak-hour congestion indices at the most affected intersections by up to 37%. These findings demonstrate that GNN-based optimization is both technically feasible and practically impactful for urban transport management in Sub-Saharan African cities.

This study presented a Graph Neural Network-Based Route Optimization and Recommendation System (GNN-RORS) for intelligent traffic prediction and route optimization within the Port Harcourt metropolitan road network. The proposed framework modelled the transportation infrastructure as a directed weighted graph in which intersections were represented as nodes and road segments were represented as dynamic edges. A hybrid Graph Attention Network and Long Short-Term Memory architecture was developed to capture both spatial traffic dependencies and temporal traffic evolution patterns, thereby enabling accurate short-term forecasting of traffic conditions and travel times. Traffic data collected from major road corridors including Ikwerre Road, Ada George Road, Government Reserved Area (GRA), and Iwofe Road were used to train, validate, and evaluate the proposed system. Experimental results demonstrated that the GAT-LSTM model significantly outperformed all benchmark approaches, including Historical Average, ARIMA, standalone LSTM, DCRNN, STGCN, and Graph WaveNet. The proposed model achieved a Mean Absolute Error of 2.34 vehicles per minute, a Root Mean Square Error of 3.17, and a Mean Absolute Percentage Error of 5.6%, confirming its ability to accurately forecast future traffic conditions within a complex urban environment. Furthermore, the route recommendation component produced substantial operational benefits. Across five representative origin-destination pairs, the system reduced average journey times from 27.2 minutes under conventional shortest-path routing to 18.7 minutes, corresponding to an average improvement of 31.2%. At the network level, congestion analysis revealed significant reductions at critical bottlenecks, particularly Rumuepirikom Junction, where congestion severity decreased by approximately 37% during peak periods. Importantly, traffic redistribution toward alternative corridors such as Iwofe Road and Ada George Road occurred without generating new congestion hotspots, demonstrating effective network-wide optimization. The findings of this study confirm that Graph Neural Networks provide an effective mechanism for modelling the complex spatio-temporal characteristics of urban transportation systems. By learning both direct and indirect traffic dependencies across interconnected road corridors, the proposed framework enables proactive route optimization rather than reactive congestion management. This capability is particularly valuable in rapidly urbanizing cities where transportation demand continues to exceed infrastructure capacity. The study therefore concludes that the proposed GNN-RORS framework is technically feasible, computationally effective, and practically beneficial for intelligent urban traffic management. Its deployment has the potential to improve commuter mobility, reduce travel delays, lower fuel consumption, minimize environmental impacts, and enhance overall transportation efficiency within Port Harcourt and other rapidly growing cities in Sub-Saharan Africa. Future work should focus on integrating real-time traffic sensors, GPS trajectory streams, incident detection mechanisms, and reinforcement learning-based adaptive routing strategies to further improve system responsiveness and scalability in real-world operational environments.

[1] S. Oyebamiji, R. Hayangah, M. Jimoh, and O. Adisa, “Sustainable Urban Mobility and Infrastructure in Sub-Saharan Africa: Engineering Pathways to Climate Mitigation and Inclusive Development,” Journal of Inclusive Cities and Built Environment, vol. 5, pp. 55-64, 2025. [2] H. B. Faheem, A. M. E. Shorbagy, and M. E. Gabr, “Impact of traffic congestion on transportation system: Challenges and remediations-a review,” Mansoura Engineering Journal, vol. 49, p. 18, 2024. [3] K. Bayon, J. Ceñido, and P. Camaro, “Analysis on the Impact of Transportation on Labor Productivity in Metropolitan Business Hubs in the Philippines,” International Journal of Research Publication and Reviews, doi, vol. 10, 2023. [4] M. A. Fattah, S. R. Morshed, and A.-A. Kafy, “Insights into the socio-economic impacts of traffic congestion in the port and industrial areas of Chittagong city, Bangladesh,” Transportation Engineering, vol. 9, p. 100122, 2022. [5] Y. Benmessaoud, L. Cherrat, and M. Ezziyyani, “Real-time self-adaptive traffic management system for optimal vehicular navigation in modern cities,” Computers, vol. 12, p. 80, 2023. [6] S. Dindar, “A Systematic Review of Urban Regeneration’s Impact on Sustainable Transport: Traffic Dynamics, Policy Responses, and Environmental Implications,” Sustainable Development, vol. 33, pp. 399-426, 2025. [7] R. Ravish and S. R. Swamy, “Intelligent traffic management: A review of challenges, solutions, and future perspectives,” Transport and Telecommunication, vol. 22, pp. 163-182, 2021. [8] E. Binshaflout, A. Hamrouni, and H. Ghazzai, “Graph Neural Networks for Vehicular Social Networks: Trends, Challenges, and Opportunities,” IEEE Transactions on Intelligent Transportation Systems, 2026. [9] Y. Li, C. Xue, F. Zargari, and Y. R. Li, “From graph theory to graph neural networks (GNNs): The opportunities of GNNs in power electronics,” IEEE access, vol. 11, pp. 145067-145084, 2023. [10] J. Xue, R. Tan, J. Ma, and S. V. Ukkusuri, “Data science in transportation networks with graph neural networks: a review and outlook,” Data Science for Transportation, vol. 7, p. 10, 2025. [11] M. Gamal and O. A. Ibrahim, “Graph neural networks for real-time optimization of autonomous urban transit systems,” Discover Cities, vol. 2, p. 102, 2025. [12] H. M. Hadi and I. M. Ibrahim, “A Comprehensive Review of Shortest Path Algorithms for Network Routing,” Asian Journal of Research in Computer Science, vol. 18, pp. 152-175, 2025. [13] A. Louati, R. Lahyani, A. Aldaej, R. Mellouli, and M. Nusir, “Mixed integer linear programming models to solve a real-life vehicle routing problem with pickup and delivery,” Applied Sciences, vol. 11, p. 9551, 2021. [14] N. G. Polson and V. O. Sokolov, “Deep learning for short-term traffic flow prediction,” Transportation Research Part C: Emerging Technologies, vol. 79, pp. 1-17, 2017. [15] J. Zhang, Y. Zheng, and D. Qi, “Deep spatio-temporal residual networks for citywide crowd flows prediction,” in Proceedings of the AAAI conference on artificial intelligence, 2017. [16] Y. Li, R. Yu, C. Shahabi, and Y. Liu, “Diffusion convolutional recurrent neural network: Data-driven traffic forecasting,” arXiv preprint arXiv:1707.01926, 2017. [17] B. Yu, H. Yin, and Z. Zhu, “Spatio-temporal graph convolutional networks: A deep learning framework for traffic forecasting,” arXiv preprint arXiv:1709.04875, 2017. [18] Z. Wu, S. Pan, G. Long, J. Jiang, X. Chang, and C. Zhang, “Connecting the dots: Multivariate time series forecasting with graph neural networks,” in Proceedings of the 26th ACM SIGKDD international conference on knowledge discovery & data mining, 2020, pp. 753-763. [19] W. Zhong, H. Meidani, and J. Macfarlane, “Attention-based spatial-temporal graph neural ode for traffic prediction,” arXiv preprint arXiv:2305.00985, 2023. [20] E. A. Aldhahri, A. A. Almazroi, M. H. Alkinani, M. Alqarni, E. A. Alghamdi, and N. Ayub, “GNN-RMNet: Leveraging graph neural networks and GPS analytics for driver behavior and route optimization in logistics,” Plos one, vol. 20, p. e0328899, 2025. [21] X. Xing, Y. Peng, L. Wan, and F. Luo, “Optimization of two-passenger ride-pooling orders based on ST-GNN and path optimization,” PLoS One, vol. 20, p. e0337415, 2025. [22] A. T. Bassey, I. N. Davies, and A. I. Akintunde, “Comparative Analysis of AI Techniques in a Traffic Advisory Model: GNN, FNN, and A*,” International Journal of Mathematics and Computer Research, vol. 13, pp. 5576-5581, 2025. [23] E. O. Iwo and S. E. Egop, “Evaluation and Prediction of Vehicular Flow in Port Harcourt,” Journal of Transportation Engineering and Traffic Management, vol. 6, pp. 33-38, 2025. [24] L. Amadi and G. Pepple, “Development of GIS-Based Road Network Map of Port Harcourt to Resolve Issues with Road Network.,” Global Science Research Journals, vol. 10, 2022. [25] F. B. C. Nwoburuoke, N. C. Bridget, E. E. Chidozie, R. A. O. Abdul, H. O. Olumide, and S. O. Akegbeyale, “An Assessment of Urban Road Transportation on Land Use Pattern in Ikwerre Road Port Harcourt,” IEEE-SEM, vol. 12, pp. 1-8, 2024.