Integrated Tracking Device & Automated Collision Detection for City Bicycles using Arduino and Global System for Mobile Communication (GSM)

Bicycle-related accidents have increased due to a lack of bike-friendly infrastructure and rising demands for city bicycles. This study aims to develop a city bicycle-mounted device that can detect collisions, determine the GPS location of the incident, and send alerts through GSM to a rescue team to increase survival by reducing emergency response time. The experimental setup detects collisions using vibration sensors, crash sensors, and accelerometer/gyroscope. Sensor data is collected during regular use and collision scenarios and stored on an SD card to determine the individual sensor thresholds for collision detection. Calculating the means of the average sensor data gathered during the crash test yields the sensor thresholds. The obtained threshold values are used in a final test involving normal use and collision scenarios. In typical use and collision scenarios, the device has 86.67 % and 90 % accuracy, respectively. Existing road characteristics such as uneven road surfaces, potholes, and the presence of litter may influence the cyclist’s control, resulting in collision detection inaccuracies. Some collisions were also undetected because the force of impact had never affected the cyclist’s control and was deemed insufficient to meet the detection threshold of the sensors.