Smart Emergency Route Optimizer System: A Systematic Literature Review

Abstract

Urban road congestion creates a serious and measurable problem for emergency medical services. Ambulances get stuck at fixed-cycle traffic signals, dispatchers rely on static maps that do not reflect current conditions, and there is rarely any mechanism to match patients to hospitals based on available treatment capability. India\'s 108 Ambulance Service records an average urban response time of around 15 minutes [A], which already exceeds the National Health Mission\'s own urban benchmark of 20 minutes [B] in many high-traffic scenarios. This paper has two goals. The first is to present a structured review of ten peer-reviewed studies from 2019 to 2025, covering AI-based dispatch, computer vision signal control, heuristic routing, autonomous vehicle frameworks, and hybrid congestion prediction. The second is to describe SEROS (Smart Emergency Route Optimizer System), a working three-tier platform built specifically for Hyderabad\'s road network, which was developed to address the gaps those studies leave open. SEROS brings together YOLOv8n for real-time traffic monitoring, A* routing with congestion-sensitive edge costs, a hospital capability scoring model, signal preemption communication, and a Kotlin Android app for field drivers. Across thirty real Hyderabad intersections and forty-six road segments, the system cut average response time from a 15 to 20 minute manual baseline down to 6 to 10 minutes. YOLOv8n detected vehicles at 97% accuracy running at twelve frames per second on standard CPU hardware, with no GPU required.

Introduction

The text discusses the development of SEROS, an AI-based emergency response optimization system designed to reduce ambulance delays in congested urban areas like Hyderabad. The study highlights that ambulance response delays significantly reduce survival chances in cardiac emergencies, mainly due to traffic congestion, non-adaptive traffic signals, static routing systems, and poor hospital selection methods. Existing research has addressed individual aspects such as machine-learning-based dispatch systems, emergency vehicle detection using computer vision, congestion-aware routing, and hospital matching, but no integrated end-to-end system had previously combined all these features into a practical deployment.

The literature survey reviews multiple studies on ambulance routing, traffic signal preemption, AI-based congestion detection, and hospital-aware navigation. These studies contributed technologies like predictive GPS synchronization, CNN-based traffic analysis, YOLO-powered emergency vehicle detection, congestion-aware A* routing, and hospital capability scoring. However, major gaps remained, including the lack of integration between vision systems and routing engines, absence of driver navigation interfaces, overreliance on distance-only hospital selection, limited validation on Indian road networks, and poor performance under adverse weather conditions.

To address these gaps, SEROS was designed as a complete smart ambulance management platform. The system integrates:

• AI-based traffic monitoring using YOLOv8n,
• real-time congestion-aware routing,
• adaptive signal preemption,
• hospital capability matching,
• and an Android-based driver navigation application.

Conclusion

The ten papers reviewed here collectively address most of the hard sub-problems in intelligent emergency vehicle routing. Digital twin synchronisation, ML-based dispatch, camera-driven signal control, capability-weighted hospital routing, congestion-aware path selection, theoretical frameworks for criticality-aware routing, multi-constraint patient-condition modelling, hybrid congestion prediction, and proactive anomaly detection all represent genuine advances. The gap they leave collectively is integration: none of these works delivered a tested end-to-end system with a driver interface. SEROS addresses that gap for Hyderabad. The polling-based GPS strategy came from the synchronisation analysis in . Time-of-day routing multipliers substitute for the ML demand model in without requiring model maintenance. YOLOv8n connects the visual detection pipeline directly to the routing graph, closing the gap between and that no reviewed paper addressed. Capability-weighted hospital scoring builds on. Exponential congestion penalties extend the logic of. The WebSocket dispatcher mechanism implements the criticality-aware communication priority argued in. Multi-variable edge costs respond to the argument in. Continuous graph weight updates solve the offline model problem in. The spike-triggered rerouting logic mirrors the anomaly detection approach of. And the Android driver application, which is absent from every reviewed system, closes the last gap between dispatch intelligence and field execution. Three things remain unresolved. Weather-robust vision is a field-wide open problem with no validated solution in any reviewed paper. Live hospital resource integration requires standardised hospital information system APIs that do not currently exist in most Indian cities. And city-scale validation requires a larger road graph than the thirty-intersection prototype tested here. These are the natural next steps for anyone building on what this work demonstrates.

References

[1] GVK Emergency Management and Research Institute (GVK EMRI). Operational data for 108 Ambulance Service — urban response time of approximately 15 minutes, as cited in: \'Calling 108: How One Institution Pioneered Emergency Medical Services in India,\' The Better India, July 2019. https://thebetterindia.com/187773/gvk-emri-108-ambulance-emergency-healthcare-service-india/ [2] Gaidhane, A. M. et al. (2023). Examining district-level disparity and determinants of timeliness of emergency medical services in Maharashtra, India. Scientific Reports, 13, 21133. https://doi.org/10.1038/s41598-023-48713-1. [References NHM urban response time benchmark of 20 minutes.] [3] Al-Shareeda, S., Celik, Y., Bilgili, B., Al-Dubai, A., and Canberk, B. (2025). Accurate AI-Driven Emergency Vehicle Location Tracking in Healthcare ITS Digital Twin. arXiv preprint arXiv:2502.03396. [4] Selvan, C., Anwar, B. H., Naveen, S., and Bhanu, S. T. (2025). Ambulance route optimization in a mobile ambulance dispatch system using deep neural network (DNN). Scientific Reports, 15, 14232. https://doi.org/10.1038/s41598-025-95048-0 [5] Sanjay Rahul P., Sathya Moorthy A., and Maheswari M. (2025). AI-Based Dynamic Traffic Management System with Real-Time Detection and Priority Signal Optimization. International Journal of Advanced Research in Computer and Communication Engineering (IJARCCE), Vol. 14, Issue 5, May 2025. https://ijarcce.com/wp-content/uploads/2025/05/IJARCCE.2025.14504.pdf [6] Alruyshid, M. H., Lutfy, O. F., and Hammood, D. A. (2025). An AI-Based Traffic Light Control System Using YOLOv10 with Emergency and Public Bus Prioritization: A Case Study in Baghdad. International Journal of Transport Development and Integration, 9(3), 609-618. https://doi.org/10.18280/ijtdi.090314 [7] Almalki, M., Aldhahri, E., and Aljojo, N. (2025). Ambulance routing optimization based on emergency medical service availability. Discover Artificial Intelligence, 5, 297. https://doi.org/10.1007/s44163-025-00352-3 [8] Nagamani, S. and Kumar, K. R. A. (2020). Advanced A* Algorithm with Dispersion Index for Dynamic Ambulance Routing Problem using Parallel Strategies. International Journal on Emerging Technologies, 11(5), 08-16. ISSN: 0975-8364. [9] Humagain, S. and Sinha, R. (2019). Routing Autonomous Emergency Vehicles in Smart Cities Using Real Time Systems Analogy: A Conceptual Model. 2019 IEEE 17th International Conference on Industrial Informatics (INDIN), Helsinki, Finland, pp. 1097-1102. https://doi.org/10.1109/INDIN41052.2019.8972218 [10] Rabbani, M., Oladzad-Abbasabady, N., and Akbarian-Saravi, N. (2022). Ambulance routing in disaster response considering variable patient condition: NSGA-II and MOPSO algorithms. Journal of Industrial and Management Optimization, 18(2), 1035-1062. https://doi.org/10.3934/jimo.2021007 [11] Charef, A., Jarir, Z., and Quafafou, M. (2022). Smart System for Emergency Traffic Recommendations: Urban Ambulance Mobility. International Journal of Advanced Computer Science and Applications, 13(10). https://doi.org/10.14569/IJACSA.2022.0131005 [12] Nirmalkar, S., Patel, S., Kanwer, A. S., and Yadav, S. (2025). Smart Ambulance Route Optimization System Using Real-Time GPS, Dynamic Traffic Signal Control, and Google API. International Research Journal of Engineering and Technology (IRJET), 12(4), 102-108. ISSN: 2395-0056. [13] H. Dia, “An agent-based approach to modelling driver route choice behavior under the influence of real-time information,” Transportation Research Part C: Emerging Technologies, vol. 10, no. 5–6, pp. 331–349, 2002. [14] M. Treiber, A. Hennecke, and D. Helbing, “Congested traffic states in empirical observations and microscopic simulations,” Physical Review E, vol. 62, no. 2, pp. 1805–1824, 2000. [15] K. Ashok and M. E. Ben-Akiva, “Dynamic network traffic estimators and predictors: State-of-the-art review,” Transportation Research Part A, vol. 34, no. 7, pp. 573–592, 2000. [16] A. Doshi and M. M. Trivedi, “Tactical driver behavior prediction and intent inference: A review,” IEEE Intelligent Transportation Systems Magazine, vol. 3, no. 3, pp. 11–25, 2011. [17] A. Ghosh, S. Banerjee, and N. Dey, “Deep learning for intelligent traffic management: A comprehensive review,” IEEE Access, vol. 8, pp. 179–194, 2020. [18] J. Redmon and A. Farhadi, “YOLOv3: An incremental improvement,” arXiv:1804.02767, 2018. [19] M. Abadi et al., “TensorFlow: Large-scale machine learning on heterogeneous distributed systems,” arXiv:1603.04467, 2016. [20] T. Litman, “Evaluating public transportation health benefits,” Victoria Transport Policy Institute, 2010. [21] P. Varaiya, “Smart cars on smart roads: Problems of control,” IEEE Transactions on Automatic Control, vol. 38, no. 2, pp. 195–207, 1993. [22] R. Bishop, “Intelligent vehicle applications worldwide,” IEEE Intelligent Transportation Systems Magazine, vol. 1, no. 1, pp. 4–9, 2008.

Copyright

Copyright © 2026 Mrs. H Meenal, Mohd Luqmaan Mohiuddin, Rajat Shinde, Mohammed Lateef Uddin. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.