Design and Development of an Autonomous Railway Track Inspection Robot using Deep Learning and GPS Localization

Design and Development of an Autonomous Railway Track Inspection Robot using Deep Learning and GPS Localization

ManuKiran KK (1), N Subhash Chandra Bose (1), Nagaraju D (1), Nagendra M (1), T M Devegowda (2), Dr. Gurupavan H R (3)

(1) UG Students, Mechanical Engineering department P.E.S. College of Engineering, Mandya Karnataka

(2) Asstistance Professor, Mechanical Engineering department P.E.S. College of Engineering, Mandya Karnataka

(3) Associate Professor, Robotics and Artificial Intelligence P.E.S. College of Engineering, Mandya Karnataka

Abstract – Railway track inspection is essential for ensuring the safety and reliability of railway transportation systems. This paper presents the design and development of an autonomous railway track crack inspection robot capable of detecting structural defects in railway tracks. The proposed system uses an ESP32- CAM module to capture real-time images of the railway track during operation.

The robot moves automatically along the track using 12 V geared DC motors, while magnetic wheels provide strong adhesion and stable movement on the metallic rails. The captured images are analyzed using image processing techniques to identify cracks and abnormalities on the track surface. When a crack is detected, the system sends an instant notification to the user through the Telegram platform for immediate action. The proposed system reduces the need for manual inspection and improves the efficiency and reliability of railway track monitoring. The system provides a low-cost and effective solution for early detection of railway track defects, thereby enhancing railway safety.

Keywords Railway Track Inspection, Crack Detection, ESP32- CAM, Autonomous Robot, Image Processing, Railway Safety.

1. INTRODUCTION

   The rapid advancement of robotics, embedded systems, and artificial intelligence has significantly improved infrastructure inspection and monitoring systems. Railway transportation is one of the most widely used modes of transport, and maintaining the structural integrity of railway tracks is essential to ensure safe and reliable operations. Cracks or structural defects in railway tracks can lead to serious accidents if they are not detected at an early stage. Traditional railway track inspection methods mainly rely on manual inspection, which is time-consuming, labor-intensive, and prone to human error. In addition, manual inspection may not always be effective in detecting small cracks or defects, especially over long railway networks.

   To address these challenges, automated robotic inspection systems have emerged as an efficient and reliable solution for monitoring railway infrastructure. These systems utilize robotic platforms equipped with cameras, sensors, and intelligent processing techniques to inspect railway tracks with higher accuracy and efficiency. By automating the inspection process, such systems reduce human involvement in hazardous environments and enable continuous monitoring of track conditions.

   In this work, an Autonomous Railway Track Crack Inspection Robot is proposed for detecting cracks and defects in railway tracks. The robot is equipped with an ESP32-CAM module that captures real-time images of the railway track surface during movement. The robot moves autonomously along the railway rails using 12 V geared DC motors, while magnetic wheels ensure stable movement and firm adhesion to the metallic rails. The captured visual data is analyzed using image processing techniques to identify cracks and structural abnormalities on the track surface.

   When a crack or defect is detected, the system generates an immediate alert and sends a notification to the user through the Telegram platform, enabling timely maintenance and preventive action. The proposed system aims to improve inspection efficiency, reduce dependence on manual monitoring, and enhance railway safety by enabling early detection of track defects. This robotic inspection approach provides a cost-effective and reliable solution for modern railway infrastructure monitoring.

2. LITERATURE REVIEW

   Zeinab Sam Daliril, et al. [1] The authors have proposed a railway security system that integrates wireless sensor networks with fuzzy logic to enable real-time monitoring and anomaly detection across railway infrastructure. Their approach focuses

   on enhancing safety and reliability by continuously analyzing data to identify potential threats or irregularities. The system allows for quick response and preventive maintenance, thereby minimizing the chances of accidents and operational failures. This research highlights how intelligent sensing and decision- making can strengthen railway safety management through automation.

   Sireesha R, et al. [2] The authors have developed a broken rail detection system that utilizes Radio Frequency (RF) technology to identify faults or discontinuities in railway tracks. The proposed system employs sensors that continuously monitor track integrity and send instant alerts to control centers upon detecting any abnormality. This enables immediate corrective action, reducing delays and preventing possible derailments. Their method ensures efficient fault detection and improved passenger safety, representing a cost-effective solution for modern railway networks.

   P. S. Bharti, et al. [3] The authors have developed a sophisticated bridge crack inspection system focused on minimizing structural risks through early detection and automated analysis of cracks. Their system utilizes unmanned aerial vehicles equipped with high-resolution cameras to collect comprehensive images of bridge surfaces, making inspections safer and more efficient. Advanced artificial intelligence algorithmsincluding specialized convolutional neural networksprocess these images in real time, precisely locating and classifying various types of cracks. This integrated method enables accurate mapping and continuous monitoring of defect- prone areas, which supports timely maintenance and proactive interventions to prevent major structural failures. The research underscores the significance of smart inspection technologies and predictive maintenance models in enhancing bridge safety, extending infrastructure lifespan, and ensuring reliable transportation networks through automated, data-driven asset management strategies.

   S. Ramesh, et al. [4] The authors have proposed an automated railway track crack detection system using ultrasonic sensors combined with embedded microcontroller technology. Their system works by transmitting ultrasonic waves along the track surface and analyzing the reflected signals to detect cracks or discontinuities. The integration of real-time data processing enables immediate identification of faults, even in harsh environmental conditions. The system also includes a wireless communication module to transmit alerts to nearby stations or maintenance teams. This approach significantly reduces manual inspection efforts, enhances detection accuracy, and ensures timely maintenance, thereby improving railway safety and operational efficiency.

   M. A. Khan, et al. [5] The authors have developed an IoT- based railway track monitoring system designed for continuous and remote supervision of track conditions. The system uses a network of sensors, including vibration, temperature, and strain sensors, to collect real-time data from railway tracks. This data is transmitted to a cloud platform where it is analyzed using data analytics techniques to detect anomalies and predict potential failures. The system supports predictive maintenance by identifying early signs of track degradation, tus preventing accidents and reducing maintenance costs. Their work emphasizes the role of IoT and cloud computing in transforming traditional railway monitoring into a smart and efficient system.

   SR. Sharma, et al. [6] The authors have introduced a computer vision-based railway track inspection system utilizing image processing techniques for crack detection. High-resolution cameras mounted on moving inspection units capture continuous images of railway tracks. These images are processed using edge detection, segmentation, and pattern recognition algorithms to identify cracks and structural defects. The system offers high detection accuracy and minimizes human error by automating the inspection process. Additionally, it supports real-time monitoring and reporting, enabling faster decision-making for maintenance operations. This research demonstrates the effectiveness of vision-based systems in enhancing railway safety through automated inspection technologies.

   R. K. Singh, et al. [7] The authors have proposed a smart railway track monitoring system based on machine learning techniques for predictive fault detection. Their system collects data from multiple sensors such as accelerometers, gyroscopes, and acoustic sensors installed along the railway tracks. This data is processed using machine learning algorithms to identify patterns associated with track defects, including cracks, misalignments, and wear. The system continuously learns from incoming data, improving its detection accuracy over time. It also provides early warning alerts to maintenance teams, enabling proactive repairs. This approach enhances the efficiency of railway maintenance operations and significantly reduces the risk of accidents by leveraging intelligent data- driven decision-making.

   A. Verma, et al. [8] The authors have developed a GPS and GSM-based railway track fault detection and alert system aimed at improving real-time communication and location tracking. The system integrates sensors to detect cracks and abnormalities in railway tracks, while the GPS module identifies the exact location of the fault. The detected information is transmitted via GSM technology to railway authorities and control centers. This ensures quick

   identification and response to potential hazards. The system is cost-effective, easy to implement, and suitable for large-scale deployment across railway networks. Their research highlights the importance of combining location tracking and wireless communication technologies for efficient fault management and improved railway safety.

   This project focuses on developing an intelligent and automated solution for railway track monitoring using embedded systems and image processing techniques. The system aims to enhance safety, reduce manual effort, and enable real-time detection of structural defects.

3. EXPERIMENTAL SETUP

   The project begins with the design and fabrication of the chassis, which is constructed using mild steel plates and rectangular sections cut precisely to the required dimensions to ensure structural strength and proper clearance for all components. Following this, the magnetic wheel assembly is developed by mounting two Ø100 mm magnetic wheels on either side of the chassis using shafts and bearings, ensuring stable alignment and strong adhesion to the railway track. Next, motor mounting is carried out by fixing 12V geared DC motors onto the chassis brackets, with proper shaft alignment to enable efficient torque transmission to the wheels. The camera and servo system is then installed at the front section of the robot, where a servo motor is used to control the angular movement of the Raspberry Pi camera for effective crack detection. After this, electronic module integration is performed by installing the Raspberry Pi, motor driver, buck converter, relay circuits, and other components within a protective housing provided on the chassis. The battery is then securely placed to maintain proper weight distribution and reduce vibration effects. Wiring and power distribution are carried out systematically by routing all electrical connections through designated paths and dividing the power supply into 12V for motors and 5V for control units using a buck converter. Once all subsystems are integrated, the final assembly is completed by aligning all components according to the design specifications, ensuring proper positioning of wheels, camera, and structural balance. Finally, the system undergoes testing and evaluation to verify its performance in detecting cracks and maintaining stability on the track, and necessary modifications are implemented to improve efficiency, accuracy, and reliability of the overall system.

   The flowchart presents a step-by-step methodology for developing the proposed system, starting with problem identification and followed by planning and design. It includes material selection, fabrication, and mechanical assembly to build the physical structure. Electronics integration and programming are then carried out to enable system functionality and control. Finally, the system is assembled, tested, and

   evaluated, with necessary modifications made to improve performance and reliability.

   FIG..1: WORKING FLOWCHART.

   1. COMPONENTS

      The below mentioned components are used to design and develop the railway track inspection robot. Each component plays a crucial role in ensuring efficient operation, control, and defect detection.

      • Raspberry Pi

      • ESP32-CAM Module

      • HC-05 Bluetooth Module

      • 12V 60 RPM Geared DC Motors

      • Servo Motor

      • Motor Driver Module / Relay Module

      • LiPo Battery (12V)

      • Magnetic Wheels

      • Mild Steel Sheet (Chassis Material)

      • Connecting Wires and Mounting Hardware

        The railway track crack inspection robot operates through the coordinated interaction of mechanical, electronic, and intelligent image-processing systems. The entire system is powered by a 12V battery, which supplies energy to the DC motors, Raspberry Pi, camera module, and auxiliary components through a regulated power distribution system. A buck converter is used to step down the voltage to 5V for safe operation of the Raspberry Pi and associated electronics.

        Robot movement is achieved using magnetic wheels driven by 12V geared DC motors, ensuring strong adhesion and stable motion along the railway track. The motors are controlled through a motor driver circuit interfaced with the Raspberry Pi, enabling precise movement and directional control. The central control unit, Raspberry Pi, manages all operations including motor control, image processing, and alert generation.

        The inspection system utilizes a Raspberry Pi Camera Module mounted at the front of the robot on an adjustable mechanism. As the robot moves, the camera continuously captures real-time images of the railway track surface. These images are processed by a pre-trained crack detection model running on the Raspberry Pi, which analyzes the track for defects, discontinuities, or cracks.

        An alert system is integrated using a buzzer connected to the GPIO pins of the Raspberry Pi. When a crack is detected, the system immediately activates the buzzer to provide an audible warning, enabling quick identification of faulty track sections.

        All subsystems work together to ensure efficient monitoring of railway tracks, enabling real-time crack detection, rapid alert generation, and improved safety through timely maintenance intervention.

   2. DESIGNING OF RAILWAY TRACK INSPECTION

      ROBOT

      This section describes the detailed CAD-based design and mechanical configuration of the railway track crack detection robot. It focuses on the strctural framework, magnetic wheel mechanism, camera inspection system, and electronic integration, all combined into a compact and efficient unit. Each subsystemincluding the chassis structure, magnetic wheel assembly, motor drive system, camera and servo mechanism, and electronic control moduleis designed with proper dimensional accuracy, material selection, and functional requirements. Special attention is given to ensuring stability on railway tracks, precise alignment of components, and effective crack detection capability. The complete CAD assembly

      illustrates the spatial arrangement, component positioning, and interaction between mechanical and electronic systems, providing a clear virtual representation for analysis, fabrication, and performance evaluation of the proposed system.

      FIG.3.2: 2D Design

      FIG.3.3: Parts of Railway Track Inspection Robot

      Assembly

      The complete CAD assembly of the railway track crack detection robot begins with the development of the base chassis model using precise dimensions and material specifications to ensure structural stability and proper component accommodation. Individual components such as magnetic wheels, shafts, DC motors, camera mount, servo bracket, and electronic housing are modeled separately in a 3D CAD environment using Solid Edge. The magnetic wheel assembly and motor drive system are designed as subassemblies to ensure accurate alignment and efficient power transmission. The camera and servo mechanism is also developed as a separate subassembly to enable controlled angular positioning for effective crack detection. These subassemblies are then integrated into the main assembly using appropriate constraints such as axial alignment, planar mating, and fixed positioning to maintain proper orientation and stability. Adequate spacing and

      clearances are provided between components to avoid interference and ensure smooth operation of moving parts. The electronic components, including the Raspberry Pi, battery, motor driver, and wiring layout, are positioned within a dedicated enclosure on the chassis, with routing paths defined using reference sketches. Exploded views are generated to illustrate the internal arrangement and assembly sequence of all components, while isometric, front, and side views are created for detailed documentation. The final CAD assembly offers a complete virtual representation of the system, enabling accurate visualization, analysis, and validation prior to fabrication.

      FIG.3.4: 3D Cad Model

   3. FABRICATION

      The following steps outline the complete fabrication procedure of the railway track inspection robot.

      • Material Selection and Cutting Mild-steel hollow rectangular pipes are selected and cut to required dimensions as per the CAD design.

      • Surface Preparation Cut sections are cleaned to remove rust, dust, and burrs for proper welding and fitting.

      • Layout and Alignment All sections are arranged on a flat surface and aligned accurately using measuring tools.

      • Tack Welding Temporary welds are applied at joints to hold the structure and allow minor adjustments.

      • Final Welding Continuous welding is performed on all joints, and cross-members are added to improve rigidity.

      • Motor Mount Design and Installation Motor mounting brackets are fabricated and welded to the chassis.

      • Motor Coupling with Shaft The motor shaft is coupled with the driven shaft using a suitable coupler for power transmission.

      • Drilling, Mounting, and Functional Fittings Holes are drilled and mounts for mechanisms and accessories are installed.

      • Grinding, Painting, and Final Finishing Welded joints are ground smooth and the chassis is painted for corrosion protection.

        The figure illustrates the fabricated mild-steel chassis with integrated DC motor mounts and wheel assembly for guided rail movement. The structure provides mechanical stability, precise alignment, and a robust base for integrating electrical and inspection modules.

        FIG.3.5: Fabrication. (a)

        FIG.3.5: Fabrication. (b)

        FIG.3.5: Fabrication. (b)

   4. ELECTRONICS INTEGRATION

      The electronics integration of the Railway Track Crack Inspection Robot brings together all hardware components into a coordinated and efficient control system.The central controller manages sensing, processing, communication, and actuation to ensure smooth and automated operation.

      control should allow flexible operation and precise movement when required. The power system, supported by a 12V battery and regulated supply, is expected to ensure stable operation of all components for a sufficient duration of inspection.

      FIG.3.6: Circuit block diagram.

      • Raspberry Pi (Central Controller): Acts as the brain of the system, processing image data, controlling peripherals, and executing inspection algorithms.

      • Power Supply Unit: Provides regulated power to all electronic components, ensuring stable operation.

      • Camera Module: Captures real-time images of the railway track for crack detection and analysis.

      • GPS Module: Tracks the robots location, enabling

        mapping and identification of defect positions.

      • Motor Driver: Interfaces between the Raspberry Pi and motors, controlling speed and direction.

      • Motors (Motor1 & Motor2): Enable movement of the robot along the railway track.

      • Buzzer: Provides audible alerts when a crack or fault is detected.

      • Wireless Communication (Wi-Fi): Allows data transmission to a remote monitoring system (laptop/PC) for real-time supervision.

   5. RESULTS & DISCUSSION

The Railway Track Crack Inspection Robot is expected to perform continuous and accurate inspection of railway tracks using a Raspberry Pibased vision system, ensuring reliable detection of cracks and surface defects with minimal human intervention. The magnetic wheel mechanism is expected to provide strong adhesion and stable movement along the railway track, preventing slipping and ensuring smooth traversal even on uneven surfaces. The camera-based inspection system should capture clear, real-time images of the rail surface, allowing the crack detection model to identify defects with high accuracy and consistency.

The buzzer alert system is expected to respond instantly upon detection of cracks, providing immediate and clear audible indications to maintenance personnel. The robots automatic mode should enable uninterrupted inspection of long track sections, while the optional manual mode using Bluetooth

Overall, the system is expected to reduce manual inspection effort, improve detection speed and accuracy, minimize human error, and provide a cost-effective solution for real-time railway track monitoring and safety enhancement.

Conclusion

The developed Railway Track Crack Inspection Robot presents an efficient and compact solution for automated railway monitoring. It combines mechanical design, embedded systems, and image processing to enhance safety and reduce manual effort.

• Integrated System Design: Combines mechanical structure, embedded systems, and image processing into a single efficient unit.

• Early Crack Detection: Identifies track defects at an early stage, reducing the risk of accidents and derailments.

• Enhanced Safety: Provides immediate alerts to improve overall railway safet.

• Stable Movement Mechanism: Magnetic wheels ensure strong grip and stability on railway tracks.

• Raspberry Pi-Based Vision System: Enables intelligent and automated inspection with minimal human intervention.

• Reduced Manual Dependency: Minimizes the need for traditional manual track inspection methods.

• Cost Efficiency: Lowers operational and maintenance costs through automation.

• Improved Maintenance Planning: Allows timely detection of defects, leading to better maintenance scheduling.

• Scalable Solution: Offers potential for expansion and deployment across large railway networks.

• Reliable Infrastructure Monitoring: Contributes to building a safer and more dependable transportation system.

  Future Scope

  Future enhancements can significantly improve the efficiency, accuracy, and autonomy of the railway inspection robot. By integrating advanced technologies, the system can evolve into a fully automated, intelligent, and reliable solution for large- scale railway monitoring.

• GPS-Based Tracking & Autonomous Navigation: Enables the robot to inspect long railway routes automatically without manual control.

• IoT-Based Communication: Supports real-time data transmission to centralized monitoring systems for remote supervision and logging.

• Obstacle Detection & Path Planning: Enhances safety and adaptability in complex track environments.

• Improved Power Systems: Use of high-capacity batteries or solar-assisted power to extend operating time.

ACKNOWLEDGMENT

This project report on, is present form is not a singular effort. We would like to acknowledge the help and guidance given by our guide Prof. T M DEVEGOWDA, and Dr. GURUPAVAN H R, our H.O.D. Dr. SRINIVASA M R, for great effort and help in our project. Without their help, this project work of ours would have been an uphill task. It has been made possible through necessary guidance and help from different quarters.

REFERENCES

1. R. Dai, Crack Detection in Civil Infrastructure Using Autonomous Robotic Systems: A Review of Platforms and Technologies, IEEE Access, Vol. 13, 2025.

2. H. Kaveh, Recent Advances in Crack Detection Technologies for Structural Health Monitoring, Frontiers in Built Environment, Vol. 10, 2024.

3. J. Kim, S. Park, H. Lee, Crack Detection in Concrete Structures Using

   HumanRobot Collaboration, arXiv, 2025.

4. B. Ramalingam, R. Rajkumar, Deep Learning Based Pavement Inspection Using Self-Reconfigurable Robot, Sensors, Vol. 21, No. 4, 2021.

5. W. Rui, Y. Kawamura, An Automated Sensing System for Steel Bridge Inspection Using Magnetic Climbing Robot, International Journal of Distributed Sensor Networks, Vol. 12, No. 6, 2016.

6. A. Tseng, S. Kalaycioglu, Intelligent Magnetic Inspection Robot for Structural Health Monitoring of Ferromagnetic Infrastructure, arXiv, 2024.

7. S. Sureshkumar, C. Agash, Augmented Reality and IoT-Based Crack Detection System, IEEE International Conference on Smart Computing, 2021.

8. Development of an Artificial Intelligence Based Railway Track Crack Detection Robot, International Journal on Recent and Innovation Trends in Computing and Communication, Vol. 11, No. 3, 2023.

9. IoT Based Railway Track Crack Detection System, SSRN Electronic

Journal, 2024.