Motorcycle Crash Detection and Alert System using IoT

Introduction

Motorcycle travel, while convenient, is considered the most dangerous form of transport due to the exposure of riders to their surroundings. Unlike cars, motorcycles lack protective enclosures, making riders more vulnerable to crashes and fatal injuries. This project focuses on developing a motorcycle crash detection and alert system using IoT to reduce the response time of emergency services. By utilizing sensors like the MPU6050 accelerometer and Firebase cloud technology, the system detects crashes and sends alerts to emergency contacts, potentially saving lives by providing real-time crash location information.

Literature Survey

The concept of safety systems for motorcycles has evolved over time, with various solutions proposed in literature. Smart helmets, as discussed by Bonnells et al. (2017), aim to enhance rider safety, although the power consumption of such systems can be high. Similarly, intelligent braking systems, proposed by Kumari et al. (2020), help prevent accidents but come with higher installation and maintenance costs. In contrast, crash detection systems using accelerometers and communication modules offer a more cost-effective and scalable solution. This project builds on such approaches, using the MPU6050 sensor to detect crashes and leveraging IoT for real-time alerts.

Proposed Methods

The system is composed of multiple modules:

1. Crash Detection: The MPU6050 multi-axis accelerometer is used to detect significant changes in acceleration that indicate a crash. The sensor measures acceleration across x, y, and z axes, and a crash is detected when a specified threshold is exceeded.

2. Data Transmission: Once a crash is detected, the system sends the crash data, including the motorcycle's location, to the Firebase cloud. From there, the data is retrieved by the rider's smartphone.

3. Alert System: The smartphone sends an alert containing the crash details (location, time) to the rider's pre-registered emergency contacts. A false alert mechanism is also included, allowing the rider to cancel the alert if no actual crash occurred.

Results and Discussion

The system was tested under various crash scenarios, including forward and side crashes. The MPU6050 sensor effectively detected these crashes by measuring the acceleration changes. The data was transmitted to Firebase and retrieved by the smartphone app, which then sent out alerts within 10 seconds of crash detection. The false alert mechanism worked as intended, allowing the rider to cancel the alert within a specified time window.

The system's performance was evaluated based on its detection accuracy and response time. The results showed that the accelerometer readings provided reliable crash detection, with a 96% accuracy in differentiating between crashes and normal driving conditions. However, improvements in GPS accuracy and the alert transmission process could further enhance the system's reliability.

Conclusion and Future Enhancements

This project successfully developed a motorcycle crash detection system using IoT, providing real-time alerts to emergency contacts. The system demonstrates the potential of using low-cost sensors and cloud technologies to improve road safety for motorcyclists. Future research could focus on refining the system by improving GPS accuracy, adding more emergency contacts, and reducing false positives. Additionally, integrating the system with wearable devices could offer more comprehensive crash detection capabilities.

➜ click here to view and download