Survey on Transit Edge: Dynamic Task Management in Collaborative Rail Computing

Abstract

The growing adoption of Intelligent Transportation Systems (ITS) demands reliable real-time monitoring, predictive maintenance, and safety-critical responses in smart rail networks. However, embedded devices such as the ESP32 are constrained by limited computation, memory, and energy efficiency, restricting their suitability for large-scale deployments. To address these challenges, we propose TRANSIT EDGE, a Train–Edge–Cloud (TEC) collaborative framework that distributes sensing, computation, and decision-making tasks across train nodes, edge servers, and the cloud. Multi-sensor data—including driver health metrics, obstacle detection, fire alertsare dynamically prioritized using a Q-learning reinforcement learning (RL) scheduler, while a quantum-inspired optimization layer accelerates convergence.The framework’s modular and scalable design further ensures adaptability to future extensions such as predictive maintenance, multi-train coordination, and integration with next-generation 5G/6G communication networks. These results establish TRANSIT EDGE as a cost-effective, intelligent, and deployment-ready solution for enhancing safety, efficiency, and resilience in modern rail systems.

Introduction

???? Objective

To develop a Train–Edge–Cloud (TEC) collaborative system called TRANSIT EDGE that improves safety, responsiveness, and scalability in urban rail transit, using ESP32-based IoT hardware, multi-sensor monitoring, and AI-powered scheduling.

???? Key Features and Technologies

?? Hardware & Sensors

• ESP32 Microcontrollers: Cost-effective, low-power edge devices.

• BPM108: Tracks driver’s temperature & blood pressure.

• Ultrasonic Sensors: Detect trackside obstacles.

• Flame Sensors: Detect fire hazards.

• ESP32-CAM: Real-time video streaming for driver authentication.

• Emergency Button: Manual hazard alert.

???? Architecture (TEC Model)

• Train Layer: Sensor data acquisition and preliminary authentication.

• Edge Layer: Task classification, Reinforcement Learning (RL)–based scheduling, quantum-inspired optimization.

• Cloud Layer: Predictive analytics, historical logging, Firebase alerts, and remote dashboard (via MIT App Inventor).

???? System Performance Highlights

???? Comparative Analysis & Literature Review

TRANSIT EDGE advances beyond these by:

• Combining multi-sensor integration

• Embedding quantum-optimized RL

• Real-world deployment on ESP32 hardware

• Enabling real-time alerts and analytics

???? Identified Challenges

1. Sensor False Alarms: Especially under rain or reflections.

2. ESP32 Scalability: Limited under high sensor/data load.

3. Network Congestion: Causes alert delay (~350 ms under 4G).

4. RL Adaptation Lag: Slower response during highly dynamic workloads.

5. Video Streaming Interruptions: Affect driver verification.

????? Proposed Improvements

? System Strengths

• Low Latency response for critical events

• High Throughput and resource efficiency

• Hardware feasibility using ESP32

• Integrated Safety Monitoring: Health, environment, and security

• Scalable Architecture adaptable to larger rail systems

? Failure Scenarios

• Sensor conflict during simultaneous hazard triggers

• Delayed alerts due to poor network conditions

• Authentication failure from video feed interruptions

Conclusion

The proposed TRANSIT EDGE framework offers a practical and future-ready approach to enhancing railway safety and performance. By combining ESP32-based sensing, Train–Edge–Cloud (TEC) collaboration, reinforcement learning (RL), and quantum-inspired optimization, it achieves low-latency, scalable, and resource-efficient operations. Experimental validation confirmed its effectiveness, with latency reduced to 120 ms (45% faster than conventional IoT systems) and throughput reaching 95 events/min. Intelligent task distribution—65% at the edge, 25% in the cloud, and 10% locally—prevented hardware overload, while quantum-inspired optimization improved decision accuracy by 12%. The framework also attained a 92% offloading success rate, enabling continuity even under fluctuating network conditions. At the application level, TRANSIT EDGE provides sub-150 ms emergency alerts, continuous driver health monitoring using BPM108, and real-time video authentication via ESP32-CAM and Firebase. This integration establishes a layered defence mechanism, combining environmental hazard detection with human-centric safety assurance. Unlike simulation-heavy studies, TRANSIT EDGE has been validated on real hardware, confirming feasibility in dynamic railway environments. Its lightweight, cost-effective, and modular design makes it suitable for wide-scale adoption, particularly in developing regions. Moreover, it is adaptable to future advancements such as 5G networks, predictive maintenance, and multi-train coordination. By shifting focus from infrastructure-only monitoring to a human–machine collaborative safety ecosystem, TRANSIT EDGE strengthens trust, resilience, and long-term sustainability in modern rail networks

References

[1] K. Lu, B. Han, and X. Zhou, “Smart urban transit systems: From integrated framework to interdisciplinary perspective,” Urban Rail Transit, vol. 4, no. 3, pp. 145–156, 2018. [Online]. Available: https://doi.org/10.1007/s40864-018-0088-0 [2] C. Wang, L. Wang, H. Chen, Y. Yang, and Y. Li, “Fault diagnosis of train network control management system based on dynamic fault tree and Bayesian network,” IEEE Access, vol. 9, pp. 12462–12472, 2021. [Online]. Available: https://doi.org/10.1109/ACCESS.2020.3048022 [3] H. Tang, Z. Yuan, H. Dai, and Y. Du, “Fault diagnosis of rolling bearing based on probability box theory and GA-SVM,” IEEE Access, vol. 8, pp. 161123–161134, 2020. [Online]. Available: https://doi.org/10.1109/ACCESS.2020.3019741 [4] L. Shang, Q. Yang, J. Wang, S. Li, and W. Lei, “Detection of rail surface defects based on CNN image recognition and classification,” in Proc. 20th Int. Conf. Adv. Commun. Technol. (ICACT), Korea, 2018, pp. 1363–1366. [Online]. Available: https://doi.org/10.23919/ICACT.2018.8323865 [5] T. de Bruin, K. Verbert, and R. Babuška, “Railway track circuit fault diagnosis using recurrent neural networks,” IEEE Trans. Neural Netw. Learn. Syst., vol. 28, no. 3, pp. 523–533, Mar. 2017. [Online]. Available: https://doi.org/10.1109/TNNLS.2016.2565660 [6] X. He, C. Xu, R. Chen, J. Wang, and J. Li, “Research on equipment health diagnosis system of railway passenger station based on cloud-edge collaboration,” J. Phys. Conf. Ser., vol. 1955, no. 1, p. 012086, 2021. [Online]. Available: https://doi.org/10.1088/1742-6596/1955/1/012086 [7] Zhao, W. Chen, Z. Cao, and L. Bai, “High-speed railway intelligent traffic control system based on cloud-edge collaboration,” IEEE Trans. Intell. Transp. Syst., 2023. [Online]. Available: https://doi.org/10.1109/TITS.2023.3275741 [8] S. Chen, Q. Li, M. Zhou, and A. Abusorrah, “Recent advances in collaborative scheduling of computing tasks in an edge computing paradigm,” Sensors, vol. 21, no. 3, p. 779, 2021. [Online]. Available: https://doi.org/10.3390/s21030779 [9] M. Chen and Y. Hao, “Task offloading for mobile edge computing in software defined ultra-dense network,” IEEE J. Sel. Areas Commun., vol. 36, no. 3, pp. 587–597, Mar. 2018. [Online]. Available: https://doi.org/10.1109/JSAC.2018.2815378 [10] K. Wang et al., “Joint offloading and charge cost minimization in mobile edge computing,” IEEE Open J. Commun. Soc., vol. 1, pp. 109–119, 2020. [Online]. Available: https://doi.org/10.1109/OJCOMS.2020.2969643 [11] S. K. U. Zaman et al., “Limpo: Lightweight mobility prediction and offloading framework using machine learning for mobile edge computing,” Cluster Comput., vol. 26, pp. 1613–1628, 2023. [Online]. Available: https://doi.org/10.1007/s10586-022-03628-2 [12] M. Mukherjee, V. Kumar, A. Lat, M. Guo, R. Matam, and Y. Lv, “Distributed deep learning-based task offloading for UAV-enabled mobile edge computing,” in Proc. IEEE INFOCOM Workshops (INFOCOM WKSHPS), Canada, 2020, pp. 1123–1128. [Online]. Available: https://doi.org/10.1109/INFOCOMWKSHPS50562.2020.91

Copyright

Copyright © 2025 Radhika B, Dr. Kamalakshi Naganna, Shilpashree B L, Tejaswini C, Vidyashree V Naik. 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.