INTRODUCTION

Indian Railways, a key player in India's logistics and trade, operates over 1.2 million kilometers of track and runs 9,000 freight trains daily from 7,349 train depots, moving 3.3 million tonnes of goods. Despite its scale, challenges such as high logistics costs and underutilization of freight capacities persist. In 2021-22, underloading costs totaled Rs. 593 crore, far exceeding loading contracts at Rs. 276 crore. Coal, comprising half of freight movements, highlights the need for efficient loading strategies. This study aims to develop a comprehensive solution addressing these challenges by monitoring loads outside designated shelters, detecting underloading, overloading, and ensuring optimal load distribution within wagons.

METHODOLOGY

The primary setup includes two load cells per axle connected to an Arduino UNO via an HX711 Amplifier for detecting load values. A secondary setup involves an ESP8266 (NodeMCU) establishing connectivity with the Arduino UNO to transmit data and establish an internet connection. Data is sent to Firebase for real-time storage and displayed on a Flutter-based mobile app, allowing users to monitor wagon loading conditions via LED indicators and take timely actions.

NodeMCU processes and outputs system data using a hybrid communication setup. One Arduino UNO communicates with NodeMCU as a master via I2C, while another Arduino UNO operates as a master communicating with NodeMCU via SPI. This approach ensures efficient data transfer from slave devices (Arduino UNO boards) to the master device (NodeMCU) without connectivity conflicts.

UNIQUE SELLING POINTS

• Versatile Load Measurement: It enables load measurement not only in loading shelters but anywhere, providing flexibility in monitoring and managing cargo weight at various points of operation.
• Axle Load Distribution Monitoring: Ensures optimal load distribution across both axles, which is crucial for extending the lifespan of axles and wheels. By preventing overloading on individual axles, it promotes better maintenance practices and reduces wear and tear.
• Affordability: Unlike competitors like PJM in Germany, whose solutions can cost up to 2 lakhs per unit, this system is significantly more cost-effective. It provides a cost-efficient alternative without compromising on functionality or reliability.
• Real-time Data Accessibility: Utilising Firebase's real-time database, the system offers instant access to data from any location with internet connectivity. This capability enhances operational efficiency by providing timely insights into load conditions and allowing prompt corrective actions.
• Easy Integration and Scalability: The integration with Arduino, HX711 load cells, NodeMCU, and Firebase ensures a straightforward setup process and scalability. Additional sensors or modules can be easily incorporated to expand functionality as needed.
• User-friendly Mobile App: The accompanying mobile application offers a user-friendly interface for monitoring and managing load data remotely. It supports customizable alerts, detailed analytics, and historical data review, empowering users to make informed decisions swiftly.

EXPERIMENTAL SETUP

The original dimensions of a wagon were scaled down by a factor of 1/10 for practical reasons. Consequently, the volume and mass of the system, with density held constant, were reduced by a factor of 1/1000. These original and scaled-down dimensions are provided by the Indian Railways website and listed in Table 1.

In our code implementation, we calibrate the system to account for the inherent weight of the board placed on the load cell, measured at 195 grams, by setting this as the baseline, effectively zeroing the load measurement. This calibration ensures that the net weight measured by the load cells excludes the board's weight. This allows for accurate determination of the variance between the standard load and the current load, considering the predefined offset.

The proposed setup monitors loading conditions on one axle of a train wagon using subsystems comprising two load cells connected to an Arduino UNO via an HX711 amplifier. Each axle has this setup, with data from both subsystems transmitted to a NodeMCU for load imbalance detection. The Arduino UNO also interfaces with LED lights for visual feedback.

During experiments, underloading is indicated when no test weight is present on the platform. When the test weight is correctly placed, the system shows a standard weight condition, calibrated to recognize this as the appropriate load. Adding extra weight in addition to the test weight results in an overloading condition, as observed in our experimental setup. Yellow LED light indicates underloading condition, Green indicates acceptable loading condition and Red indicates overloading condition.

FLUTTER MOBILE APPLICATION

The NodeMCU module (ESP8266) integrates with Firebase's Real-time Database via the ESP8266Firebase library, enabling seamless data transfer from Arduino Uno microcontrollers connected to HX711 load cells. This setup consolidates data from multiple Arduino setups on wagon axles into a centralised database, ensuring convenient access to real-time information. This integration empowers the system to provide timely insights for informed decision-making.

The mobile application complements this setup by accessing the Firebase database to display and analyse the collected data. It provides users with a user-friendly interface to monitor axle loads, detect load imbalances, and track loading conditions remotely. By leveraging Firebase's real-time capabilities, the app ensures that stakeholders can make prompt decisions based on current wagon loading statuses.

RESULT

The distribution of load between axles is assessed by analysing each axle's loading condition independently. When Axle 1 shows Standard Loading while Axle 2 indicates either Overloading, it highlights an uneven distribution of load across the axles.

The Flutter mobile app dynamically displays loading conditions, including current loads and deviations from standard loads. It calculates deviations by comparing measured loads to standard values and indicates whether the load distribution across both axles is balanced. Real-time monitoring and deviation calculations in the Flutter app facilitate prompt corrective actions, supporting operational decisions and maintenance efforts.